University of Arkansas, FayettevilleScholarWorks@UARK

Theses and Dissertations

5-2018

Phylogeny and Evolutionary Genomics of Non-Photosynthetic DiatomsAnastasiia OnyshchenkoUniversity of Arkansas, Fayetteville

Follow this and additional works at: https://scholarworks.uark.edu/etd

Part of the Bioinformatics Commons, Evolution Commons, Genomics Commons, and theMolecular Genetics Commons

This Thesis is brought to you for free and open access by ScholarWorks@UARK. It has been accepted for inclusion in Theses and Dissertations by anauthorized administrator of ScholarWorks@UARK. For more information, please contact scholar@uark.edu, ccmiddle@uark.edu.

Recommended CitationOnyshchenko, Anastasiia, "Phylogeny and Evolutionary Genomics of Non-Photosynthetic Diatoms" (2018). Theses and Dissertations.2695.https://scholarworks.uark.edu/etd/2695

Phylogeny and Evolutionary Genomics of Non-Photosynthetic Diatoms

A thesis submitted in partial fulfillment of the requirements for the degree of

Master of Science in Cell and Molecular Biology

by

Anastasiia Onyshchenko Taras Shevchenko National University of Kyiv

Bachelor of Science in Biology, 2016

May 2018 University of Arkansas

This thesis is approved for recommendation to the Graduate Council ___________________ Dr. Andrew Alverson Thesis Director ___________________ Dr. Andy Pereira Committee member ___________________ Dr. Jeffrey Lewis Committee member

Abstract

Diatoms are prolific photosynthesizers responsible for some 20% of global primary production.

In real terms, the oxygen in one of every five breaths traces back to photosynthesis by marine

diatoms. Among the tens of thousands of diatom species, a small handful of colorless diatom

species in the genus Nitzschia have lost photosynthesis altogether and rely exclusively on

extracellular organic carbon for growth. I used DNA sequence data to reconstruct the phylogeny

of this group, and found that nonphotosynthetic diatoms are monophyletic, indicating that

photosynthesis was lost just one time over the course of some 200 million years of diatom

evolution. Carbon metabolism in nonphotosynthetic diatoms, including the exact source of

carbon used by these species, has not been fully characterized. We sequenced the nuclear genome

of one species and used it to develop a comprehensive model of central carbon metabolism.

Preliminary analysis of Nitzschia metabolism showed that it generally matches to the pattern of

previously reported diatom metabolic networks. As well we found some hints regarding

Nitzschia external carbon acquisition which possibly can help to explain its heterotrophic mode

of life. Overall, this study has provided novel insights into the evolutionary origin and

metabolism of non-photosynthetic diatoms, which are unique among diatoms in their ability to

sustain their growth solely from extracellular carbon.

Acknowledgments I would like to thank Fulbright Program for providing me opportunity to complete Master’s

Degree in Cell and Molecular Biology at University of Arkansas. I am grateful to my adviser,

Andrew Alverson, for a chance to work on such a great research project and support and

opportunities he provided to me during two years of my studies. I would like to thank to all the

lab members, Teofil Nakov, Elizabeth Ruck, Kala Downey for excellent work environment,

moral and professional support and for the research they are doing without which none of my

work could happen. I’d like to thank my committee members Jeffrey Lewis and Andy Pereira for

agreeing to advise and help me during completion of my degree. I am grateful to Douglas

Rhoads (Head of CEMB Program) and David McNabb (Head of Biology Department) for letting

me be part of their programs.

I would also like to thank to my mother Tetyana and my boyfriend Lance for unconditional

emotional support whenever I needed it.

Table of Contents Introduction 1

The Diatoms 1 Loss of photosynthesis in diatoms and across the tree of life 2 Nuclear genomic insights into carbon metabolism in nonphotosynthetic diatoms 4 Rationale and significance 5 References 7

Chapter 1. A Single Loss of Photosynthesis In Diatoms 1 0 Abstract 1 0 Introduction 1 1 Materials and Methods 1 4

Collection and culturing of apochloritic Nitzschia 14 DNA extraction, PCR, and DNA sequencing 1 4 Phylogenetic analyses 1 5 Plastid genome sequencing and analysis 1 7

Results 1 7 Phylogeny of apochloritic Nitzschia 1 7 Plastid genome sequencing and analysis 19

Discussion 2 1 Monophyly of apochloritic diatoms 2 1 The ecology and biogeography of apochloritic diatoms 2 3 Plastid genome reduction in apochloritic diatoms 2 4

Conclusions 2 5 References 2 7

Chapter 2. Core carbon metabolism and characterization of a β-ketoadipate pathway inferred from the genome of a non-photosynthetic diatom (Bacillariophyta) 48

Abstract 48 Introduction 49 Materials and Methods 5 0

Collection and culturing of Nitzschia sp. 5 0 DNA and RNA extraction and sequencing 5 1 Genome and transcriptome assembly and annotation 5 1 Genome annotation 5 3 Construction of orthologous clusters 5 4 Characterization of carbon metabolism genes 5 5 Prediction of protein localization 56

Results and Discussion 56 Genome characteristics 56 Central carbon metabolism 57 A β-ketoadipate pathway in diatoms 6 1

Conclusions 6 5 References 67

Conclusions 8 4

List of tables

Chapter 1. A Single Loss of Photosynthesis In Diatoms 10 Table 1-1. Taxa and sources of DNA sequences analyzed in this study. 32 Table S1-1. Primers used to amplify and sequence cob and nad1 fragments for study taxa. 37 Table S1-2. Overlapping and adjacent genes in plastid genomes of Nitzschia sp. nitz4 and Nitzschia sp. NIES-3581. Groups of overlapping or adjacent genes are separated by empty rows with the coordinates showing the degree of overlap. All but two groups are shared between the two genomes. 38

Chapter 2. Core carbon metabolism and characterization of a β-ketoadipate pathway inferred from the genome of a non-photosynthetic diatom (Bacillariophyta) 48

Table 2-1. Nitzschia sp. Nitz4 genome annotation progress statistics 72 Table 2-2. General genome annotation statistics for analyzed diatom species 72 Table 2-3. Diatom β-ketoadipate pathway annotation for analyzed species 73

List of figures

Chapter 1. A Single Loss of Photosynthesis In Diatoms 1 0 Figure 1-1. Light micrographs of newly sequenced nonphotosynthetic Nitzschia species. Clockwise from top: Nitzschia sp. nitz2, Nitzschia sp. nitz7, Nitzschia sp. nitz8, Nitzschia sp. nitz4. 3 3 Figure 1-2. Phylogenetic trees inferred from plastid 16S (A), plastid 16S transformed into purine/pyrimidine (R/Y) coding (B), nuclear 28S d1–d2 genes for all newly sequenced taxa and data from GenBank [”lsu (new + ncbi)”, C], mitochondrial cob (D), mitochondrial nad1 (E), concatenated cob and nad1 genes [”mito (cob + nad1 )”, F], concatenated cob, nad1, and nuclear 28S d1–d2 genes for all newly sequenced taxa [”mito (cob + nad1 ) + lsu (new)”, G], and a concatenated alignment of cob, nad1, and nuclear 28S d1–d2 genes for taxa from this study and GenBank [”mito (cob + nad1 ) + lsu (new + ncbi)”, H]. For the phylogenies in panels C and H, we removed branches shorter than 0.00001 units for clarity. The full phylogenies are available in Supplementary Figure 1. Thicker black branches correspond to apochlorotic taxa. White points identify nodes with bootstrap support >70%. 3 4 Figure 1-3. Conserved synteny, gene content, and sequence in the plastid genomes of two nonphotosynthetic diatoms in the genus Nitzschia, Nitz4 (this study) and NIES-3581 (Kamikawa et al. 2015a). Unlabeled genes are shared between the two species. 3 6 Figure S1-1. Phylogenetic trees for plastid 16S (A), plastid 16S transformed into purine/pyrimidine (R/Y) coding (B), mitochondrial cob (C), nad1 (D), a densely sampled 28S d1–d2 matrix with sequences from this study and GenBank (E), cob and nad1 combined (F), and a large combined nuclear and mitochondrial gene matrix with sequences from this study and GenBank (G). In the 16S tree, Nitzschia sp. NIES-3581 is identified by its synonym, iriis04. 4 0

Chapter 2. Core carbon metabolism and characterization of a β-ketoadipate pathway inferred from the genome of a non-photosynthetic diatom (Bacillariophyta) 4 8

Figure 2-1. Protocatechuate dioxygenase genes localization on Nitzschia sp. Nitz4 scaffold s 7 8 Figure 2-2. Schematic annotation of protocatechuate branch of 𝛽-ketoadipate pathway in analyzed diatom species 79 Figure S2-1. Statistics for original genome scaffolding for range of K-mers 8 0 Figure S2-2. Statistics for organellar reads-free genome scaffolding for range of K-mers 8 1 Figure S2-3. Comprehensive mitochondrial metabolic pathways of Nitzschia sp. Nitz4 derived from genome annotation 8 2 Figure S2-4. Comprehensive plastid metabolic pathways of Nitzschia sp. Nitz4 derived from genome annotation 8 3

List of published papers Chapter 1: Onyshchenko, A., Ruck, E.C., Nakov, T. & Alverson, A.J. 2018. A single loss of photosynthesis in diatoms . bioRxiv . doi: http://dx.doi.org/10.1101/298810

Introduction

The Diatoms

Diatoms (Bacillariophyta) are a group of widely distributed unicellular algae belonging to

the chromalveolate clade, which includes photosynthetic lineages such as brown algae,

dinoflagellates, haptophytes, and non-photosynthetic groups such ciliates and apicomplexans

(Cavalier-Smith 1999) . Diatoms are ancestrally photosynthetic, and they are responsible for

roughly 20% of global primary production. They are also highly diverse, with as many as

200,000 different species (Mann and Droop 1996) . Diatoms are also known for their elaborate

and and highly intricate silica cell walls (Round, Crawford, and Mann 1990) and are classified

historically based on cell wall features and, more recently, their phylogenetic relationships (Sims,

Mann, and Medlin 2006) . For example, “centric” diatoms (Coscinodiscophyceae) are radially

symmetrical, and pennate diatoms have bipolar symmetry. Raphe-bearing pennate diatoms

(Bacillariophyceae) possess a slit for their cell wall that allows them to glide along a surface.

Diatoms are ancestrally oogamous, and this mode of reproduction is found throughout the centric

diatom lineages. The pennate diatoms are isogamous or anisogamous (Williams and Kociolek

2011) . Raphid pennates constitute the vast majority of diatom diversity, which likely reflects a

combination of life history and active motility (Nakov, Beaulieu, and Alverson 2018) .

Diatom plastids surrounded by four membranes (in contrast to green algae and land

plants, which have only two membranes), which reflects the origin of their plastids in which a

which a photosynthetic red alga cell was engulfed by a heterotrophic eukaryote. Diatom cells are

highly compartmentalized, including a periplastid compartment derived from the cytoplasm of

the ancient red algal endosymbiont (Gruber et al. 2007) . As a result, diatom genomes often

1

possess multiple isozymes related to central carbon pathways that have distinct cellular

localizations. For example, several diatoms have complete or partial glycolytic pathways

localized to all three cellular compartments (Smith, Abbriano, and Hildebrand 2012; Kroth et al.

2008) . It is clear from the very small sample of sequenced diatom genomes that diatoms are

genomically and physiologically diverse, so we expect that newly sequenced genomes are likely

to reveal many new features and unsuspected peculiarities.

Loss of photosynthesis in diatoms and across the tree of life

Loss of photosynthesis is common in nearly all major groups of photosynthetic

eukaryotes (Hadariová et al. 2018) . For example, at least five different lineages within green the

algal orders Chlamydomonodales and Chlorellales lost photosynthesis and are now obligate

parasites (Těšitel 2016) . Multiple lineages of flowering plants either completely or incompletely

dispensed with photosynthesis in favor of parasitism as well (Barkman et al. 2007; Těšitel 2016) .

Photosynthesis has been lost dozens or more times in florideophycean red algae as well. In a

phenomenon known as adelphoparasitism, these species go on to parasitize their sister species

(Goff et al. 1996; Blouin and Lane 2012) . In addition to lineage with primary green and red

plastids, species with secondary plastids have repeatedly lost photosynthesis as well. These

include euglenoids (Marin 2004) , apicomplexans ( (McFadden et al. 1996) , (Waller and

McFadden 2005) ), ciliates and dinoflagellates (Taylor, Hoppenrath, and Saldarriaga 2008) . In

most cases, it is thought that secondarily nonphotosynthetic species evolved from mixotrophic

ancestors, which in addition to photosynthesis, can also use extracellular carbon for growth.

Many pennate diatoms are mixotrophic (Hellebust and Lewin 1977) , and although it is

likely very costly to maintain these dual modes of nutrition (Raven 1997) , it may help these

2

diatoms survive through periods of low sunlight when photosynthesis alone cannot sustain cell

viability (Hellebust and Lewin 1977; Tuchman et al. 2006) . Despite recurrent loss of

photosynthesis in all major clades of photosynthetic organisms, loss of photosynthesis in diatoms

has been rare. Only one subset of 20 or so mostly undescribed free-living species in the genus

Nitzschia that have given up photosynthesis and are now obligate heterotrophs (Lewin and

Lewin 1967; Kamikawa et al. 2015) . These “apochloritic” diatoms are often found in

environments that are rich in organic carbon such as mangrove forests and decaying seaweeds

(Pringsheim 1956; Lewin and Lewin 1967; Kamikawa et al. 2015; Blackburn, Hannah, and

Rogerson 2009a) . Although few in number and limited in their taxonomic scope, these species

nevertheless encompass a broad range of morphological diversity and are worldwide in

distribution (Blackburn, Hannah, and Rogerson 2009b; Kamikawa et al. 2015) , which together

raises the possibility that loss of photosynthesis occurred more than once in this group—a

hypothesis supported by phylogenetic analyses of one plastid and one nuclear gene (Kamikawa

et al. 2015) . These data could not reject monophyly of apochloritic species either (Kamikawa et

al. 2015) .

The first part of my thesis used combined DNA datasets from the nuclear, plastid,

mitochondrial genomes, as well as expanded species sampling, to test the number of losses of

photosynthesis in this group. I found that photosynthesis was lost just one time in the common

ancestor of this one small subclade within the genus Nitzschia (Onyshchenko et al. 2018) . Thus,

this radical trophic shift from autotrophy to obligate heterotrophy represents a singular rare event

in diatoms. For this study, we also sequenced and analyzed the plastid genome of one

nonphotosynthetic diatom, Nitzschia sp. Nitz4, and compared it with a previously sequenced

3

nonphotosynthetic diatom plastid genome (Kamikawa et al. 2016) . The genomes were highly

similar in gene content, synteny and sequence—consistent with rapid genomic streamlining

following the loss of photosynthesis in their common ancestor.

Nuclear genomic insights into carbon metabolism in nonphotosynthetic diatoms

Genomic and experimental studies alike have shown that central carbon metabolism in

diatoms is highly complex and compartmentalized (Kroth et al. 2008; Smith, Abbriano, and

Hildebrand 2012; Armbrust et al. 2004; Traller et al. 2016; Kamikawa et al. 2017) . These studies

have shown that relatively few generalizations about carbon metabolism can be made across

diatom species, thereby cautioning against extrapolations from model to non-model species. This

would include species that have lost photosynthesis and are likely to have unique carbon

metabolic pathways. For the second part of my thesis, I analyzed the nuclear genome sequence of

Nitzschia sp. Nitz4 to better understand carbon metabolism in nonphotosynthetic diatoms. I was

specifically interested in identifying the mechanism of carbon carbon acquisition and forms of

organic carbon used by these species.

I provide the first characterization of a β-ketoadipate pathway in diatoms. This pathway is

involved in degradation of aromatic components derived from decaying plant material (e.g.,

lignin) or from toxic environment pollutants (Harwood and Parales 1996) . The β-ketoadipate

pathway was previously thought to be restricted exclusively to fungi and bacteria, which have

slightly different versions of the pathway. I expanded my analysis to other sequenced diatom

genomes and found that genes in the β-ketoadipate pathway were present in all sequenced

diatoms, though the pathway was incomplete in most cases. I used the genome to predict that

diatoms transport and subsequently degrade protocatechuic acid, an aromatic compound derived

4

from lignin, and that the terminal products of the pathway are direct intermediates of the Krebs

cycle. I also developed a model of central carbon metabolism in this species and compared it

photosynthetic diatoms. It appears that Nitzschia sp. Nitz4 has retained the conserved

compartmentalization pattern for carbon metabolism as photosynthetic species. Modifications

include the ability to initiate gluconeogenesis exclusively within the mitochondrion and the

absence of pyruvate metabolising enzymes found in other diatoms.

Rationale and significance

A clear understanding of the loss of photosynthesis in diatoms requires a robust

phylogenetic hypothesis, and I showed that photosynthesis was lost once in this group. This

allows us to better understand the evolutionary and ecological context of this event, and in

addition, it guides strategies for future research in this group. For example, comparative

approaches offer limited insights into events that occurred just one time.

Trophic shifts from autotrophy to heterotrophy can have profound and cascading genomic

consequences, and these have been most intensively studies in the plastid genome. I focused

most of my attention on the nuclear genome, using it to understand and model carbon

metabolism in apochloritic diatoms. I found and characterized a novel carbon pathway, a finding

that I will test further with experimental techniques. In addition, this finding raised several

questions about the origin and phylogenetic distribution of this pathway across diatoms as a

whole.

Diatoms show great promise for industrial biofuel production (Traller et al. 2016; Tanaka

et al. 2015) , and the discovery of a β-ketoadipate pathway in diatoms raises new possibilities for

biotechnology applications. Finally, nonphotosynthetic diatoms are highly amenable to

5

laboratory experimentation, making it possible to test the genome-based predictions made in this

study and establishing them as a useful model for understanding carbon metabolism in diatoms

more generally.

6

References Armbrust, E. Virginia, John A. Berges, Chris Bowler, Beverley R. Green, Diego Martinez,

Nicholas H. Putnam, Shiguo Zhou, et al. 2004. “The Genome of the Diatom Thalassiosira Pseudonana : Ecology, Evolution, and Metabolism.” Science 306 (5693): 79–86.

Barkman, Todd J., Joel R. McNeal, Seok Hong Lim, Gwen Coat, Henrietta B. Croom, Nelson D. Young, and Claude W. DePamphilis. 2007. “Mitochondrial DNA Suggests at Least 11 Origins of Parasitism in Angiosperms and Reveals Genomic Chimerism in Parasitic Plants.” BMC Evolutionary Biology 7 (1). BioMed Central: 248.

Blackburn, Michele V., Fiona Hannah, and Andrew Rogerson. 2009a. “First Account of Apochlorotic Diatoms from Mangrove Waters in Florida.” The Journal of Eukaryotic Microbiology 56 (2): 194–200.

———. 2009b. “First Account of Apochlorotic Diatoms from Intertidal Sand of a South Florida Beach.” Estuarine, Coastal and Shelf Science 84 (4): 519–26.

Blouin, Nicolas A., and Christopher E. Lane. 2012. “Red Algal Parasites: Models for a Life History Evolution That Leaves Photosynthesis behind Again and Again.” BioEssays: News and Reviews in Molecular, Cellular and Developmental Biology 34 (3): 226–35.

Cavalier-Smith, T. 1999. “Principles of Protein and Lipid Targeting in Secondary Symbiogenesis: Euglenoid, Dinoflagellate, and Sporozoan Plastid Origins and the Eukaryote Family Tree.” The Journal of Eukaryotic Microbiology 46 (4). Blackwell Publishing Ltd: 347–66.

Goff, Lynda J., Debra A. Moon, Pi Nyvall, Birgit Stache, Katrina Mangin, and Giuseppe Zuccarello. 1996. “The Evolution of Parasitism in the Red Algae: Molecular Comparisons of Adelphoparasites and Their Hosts.” Journal of Phycology 32 (2). Wiley Online Library: 297–312.

Gruber, Ansgar, Sascha Vugrinec, Franziska Hempel, Sven B. Gould, Uwe-G Maier, and Peter G. Kroth. 2007. “Protein Targeting into Complex Diatom Plastids: Functional Characterisation of a Specific Targeting Motif.” Plant Molecular Biology 64 (5): 519–30.

Hadariová, Lucia, Matej Vesteg, Vladimír Hampl, and Juraj Krajčovič. 2018. “Reductive Evolution of Chloroplasts in Non-Photosynthetic Plants, Algae and Protists.” Current Genetics 64 (2): 365–87.

Harwood, C. S., and R. E. Parales. 1996. “The Beta-Ketoadipate Pathway and the Biology of Self-Identity.” Annual Review of Microbiology 50: 553–90.

Hellebust, Johan A., and Joyce Lewin. 1977. “Heterotrophic Nutrition.” The Biology of Diatoms .

7

Kamikawa, Ryoma, Daniel Moog, Stefan Zauner, Goro Tanifuji, Ken-Ichiro Ishida, Hideaki Miyashita, Shigeki Mayama, et al. 2017. “A Non-Photosynthetic Diatom Reveals Early Steps of Reductive Evolution in Plastids.” Molecular Biology and Evolution 34 (9). academic.oup.com: 2355–66.

Kamikawa, Ryoma, Goro Tanifuji, Sohta A. Ishikawa, Ken-Ichiro Ishii, Yusei Matsuno, Naoko T. Onodera, Ken-Ichiro Ishida, et al. 2016. “Proposal of a Twin Arginine Translocator System-Mediated Constraint against Loss of ATP Synthase Genes from Nonphotosynthetic Plastid Genomes.” Molecular Biology and Evolution 33 (1). academic.oup.com: 303.

Kamikawa, Ryoma, Naoji Yubuki, Masaki Yoshida, Misaka Taira, Noriaki Nakamura, Ken-Ichiro Ishida, Brian S. Leander, et al. 2015. “Multiple Losses of Photosynthesis in Nitzschia (Bacillariophyceae).” Phycological Research 63 (1): 19–28.

Kroth, Peter G., Anthony Chiovitti, Ansgar Gruber, Veronique Martin-Jezequel, Thomas Mock, Micaela Schnitzler Parker, Michele S. Stanley, et al. 2008. “A Model for Carbohydrate Metabolism in the Diatom Phaeodactylum Tricornutum Deduced from Comparative Whole Genome Analysis.” PloS One 3 (1): e1426.

Lewin, Joyce, and R. A. Lewin. 1967. “Culture and Nutrition of Some Apochlorotic Diatoms of the Genus Nitzschia.” Microbiology 46 (3). Microbiology Society: 361–67.

Mann, D. G., and S. J. M. Droop. 1996. “Biodiversity, Biogeography and Conservation of Diatoms.” In Biogeography of Freshwater Algae , 19–32. Developments in Hydrobiology. Springer, Dordrecht.

Marin, Birger. 2004. “Origin and Fate of Chloroplasts in the Euglenoida.” Protist 155 (1). Urban & Fischer: 13–14.

McFadden, G. I., M. E. Reith, J. Munholland, and N. Lang-Unnasch. 1996. “Plastid in Human Parasites.” Nature 381 (6582). Nature Publishing Group: 482.

Nakov, Teofil, Jeremy M. Beaulieu, and Andrew J. Alverson. 2018. “Accelerated Diversification Is Related to Life History and Locomotion in a Hyperdiverse Lineage of Microbial Eukaryotes (Diatoms, Bacillariophyta).” The New Phytologist , April. https://doi.org/ 10.1111/nph.15137 .

Onyshchenko, Anastasiia, Elizabeth C. Ruck, Teofil Nakov, and Andrew J. Alverson. 2018. “A Single Loss of Photosynthesis in Diatoms.” bioRxiv . https://doi.org/ 10.1101/298810 .

Pringsheim, E. G. 1956. “Micro-Organisms from Decaying Seaweed.” Nature 178 (4531): 480–81.

Raven, J. A. 1997. “Phagotrophy in Phototrophs.” Limnology and Oceanography 42 (1): 198–205.

8

Round, F. E., R. M. Crawford, and D. G. Mann. 1990. Diatoms: Biology and Morphology of the Genera . Cambridge University Press.

Sims, Patricia A., David G. Mann, and Linda K. Medlin. 2006. “Evolution of the Diatoms: Insights from Fossil, Biological and Molecular Data.” Phycologia 45 (4). The International Phycological Society: 361–402.

Smith, Sarah R., Raffaela M. Abbriano, and Mark Hildebrand. 2012. “Comparative Analysis of Diatom Genomes Reveals Substantial Differences in the Organization of Carbon Partitioning Pathways.” Algal Research 1 (1): 2–16.

Tanaka, Tsuyoshi, Yoshiaki Maeda, Alaguraj Veluchamy, Michihiro Tanaka, Heni Abida, Eric Maréchal, Chris Bowler, et al. 2015. “Oil Accumulation by the Oleaginous Diatom Fistulifera Solaris as Revealed by the Genome and Transcriptome.” The Plant Cell 27 (1): 162–76.

Taylor, F. J. R., Mona Hoppenrath, and Juan F. Saldarriaga. 2008. “Dinoflagellate Diversity and Distribution.” Biodiversity and Conservation 17 (2): 407–18.

Těšitel, Jakub. 2016. “Functional Biology of Parasitic Plants: A Review.” Plant Ecology and Evolution 149 (1): 5–20.

Traller, Jesse C., Shawn J. Cokus, David A. Lopez, Olga Gaidarenko, Sarah R. Smith, John P. McCrow, Sean D. Gallaher, et al. 2016. “Genome and Methylome of the Oleaginous Diatom Cyclotella Cryptica Reveal Genetic Flexibility toward a High Lipid Phenotype.” Biotechnology for Biofuels 9 (November): 258.

Tuchman, Nancy C., Marc A. Schollett, Steven T. Rier, and Pamela Geddes. 2006. “Differential Heterotrophic Utilization of Organic Compounds by Diatoms and Bacteria under Light and Dark Conditions.” Hydrobiologia 561 (1). Kluwer Academic Publishers: 167–77.

Waller, Ross F., and Geoffrey I. McFadden. 2005. “The Apicoplast: A Review of the Derived Plastid of Apicomplexan Parasites.” Current Issues in Molecular Biology 7 (1): 57–79.

Williams, David M., and J. Patrick Kociolek. 2011. “An Overview of Diatom Classification with Some Prospects for the Future.” In The Diatom World , edited by Joseph Seckbach and Patrick Kociolek, 47–91. Dordrecht: Springer Netherlands.

9

Chapter 1. A Single Loss of Photosynthesis In Diatoms

Abstract

Loss of photosynthesis is a common and often repeated trajectory in nearly all major groups of

photosynthetic eukaryotes. One small subset of ‘apochloritic’ diatoms in the genus Nitzschia

have lost their ability to photosynthesize and require extracellular carbon for growth. Similar to

other secondarily nonphotosynthetic taxa, apochloritic diatoms maintain colorless plastids with

highly reduced plastid genomes. Although the narrow taxonomic breadth of apochloritic diatoms

suggests a single loss of photosynthesis in their common ancestor, previous phylogenetic

analyses suggested that photosynthesis was lost multiple times. We sequenced phylogenetic

markers from the nuclear and mitochondrial genomes for a broad set of taxa and found that the

best trees for data sets representing all three genetic compartments supported monophyly of

apochloritic Nitzschia , consistent with a single loss of photosynthesis in diatoms. We sequenced

the plastid genome of one apochloritic species and found that it was highly similar in all respects

to the plastid genome of another apochloritic Nitzschia species, indicating that streamlining of

the plastid genome had completed prior to the split of these two species. Finally, it is increasingly

clear that some locales host relatively large numbers apochloritic Nitzschia species that span the

phylogenetic diversity of the group, indicating that these species co-exist because of resource

abundance or resource partitioning in ecologically favorable habitats. A better understanding of

the phylogeny and ecology of this group, together with emerging genomic resources, will help

identify the factors that have driven and maintained the loss of photosynthesis in this group, a

rare event in diatoms.

10

Introduction

Photosynthetic eukaryotes (Archaeplastida) trace back to a single common ancestor, in

which an eukaryotic host paired with a cyanobacterial endosymbiont that would eventually

become the plastid, a fully integrated cellular organelle that is the site of photosynthesis

(Archibald 2009, Keeling 2010) . Although most archaeplastids remain photoautotrophic, loss of

photosynthesis has occurred —often repeatedly—in nearly all major archaeplastid lineages

(Hadariová et al. 2017) . As many as five different lineages within the green algal orders

Chlamydomonadales and Chlorellales have traded off photosynthesis for trophic strategies that

include heterotrophy and obligate parasitism (Rumpf et al. 1996, Tartar and Boucias 2004, Yan et

al. 2015) . Partial or complete loss of photosynthesis has occurred in at least 11 different lineages

of flowering plants, representing hundreds of species (Barkman et al. 2007) ; parasitic

angiosperms obtain extracellular carbon from the vascular tissues of host plants for all

(holoparasites) or part (hemiparasites) of their life history (Těšitel 2016) . Perhaps most

strikingly, photosynthesis has been lost dozens or more times across the florideophycean red

algae (e.g., Goff et al. 1996, Kurihara et al. 2010) , which go on to parasitize a closely related

photosynthetic species—a phenomenon known as adelphoparasitism (Blouin and Lane 2012) .

Photosynthesis has been lost in a broad range of taxa with secondary plastids as well, including

euglenoids (Marin 2004) , apicocomplexans (McFadden et al. 1996, Waller and McFadden 2005) ,

ciliates (Reyes-Prieto et al. 2008) , dinoflagellates (Saldarriaga et al. 2001) , cryptophytes

(Donaher et al. 2010, Martin-Cereceda et al. 2010) , and stramenopiles (Tyler et al. 2006) . Most

nonphotosynthetic algae evolved from mixotrophic ancestors (Figueroa-Martinez et al. 2015) ,

11

likely because they already had the means to secure extracellular carbon and because the

energetic costs of mixotrophy are thought to be high (Raven 1997) .

Diatoms are a lineage of stramenopile algae responsible for roughly 20% of global

primary production (Field et al. 1998) . They are ancestrally photosynthetic, and although the

overwhelming majority of the estimated 100,000 or so diatom species remain photosynthetic,

many species are mixotrophic, which allows them to use external sources of carbon for growth in

fluctuating light conditions (Hellebust and Lewin 1977, Tuchman et al. 2006) . A much smaller

set of 20 or so mostly undescribed, colorless, free-living species in the genus Nitzschia , and one

species in the closely related and morphologically similar genus Hantzschia , have abandoned

photosynthesis altogether and rely exclusively on extracellular carbon for growth (Lewin and

Lewin 1967) . These “apochloritic” diatoms—the only known nonphotosynthetic diatom

species—are often found in association with mangroves, and decaying seaweeds and sea grasses

(Pringsheim 1956, Blackburn et al. 2009a, Kamikawa et al. 2015b) .

The small number of species and narrow taxonomic range of apochloritic diatoms leads

to the prediction, based on parsimony, that nonphotosynthetic diatoms are monophyletic, tracing

back to a single loss of photosynthesis in their common ancestor. Despite the small number of

species, however, they encompass a relatively broad range of morphological diversity

(Blackburn et al. 2009b, Kamikawa et al. 2015b) and use a variety of carbon sources (Lewin and

Lewin 1967, Hellebust and Lewin 1977) , raising the possibility that obligate heterotrophy

evolved multiple times—a hypothesis supported by a phylogenetic analysis of nuclear 28S d1–d2

sequences that separated apochloritic taxa into three separate clades (Kamikawa et al. 2015b) .

Monophyly of apochloritic species could not be rejected, however (Kamikawa et al. 2015b) .

12

These two competing hypotheses (one vs. multiple origins) have important implications for our

understanding of the underlying phylogenetic, ecological, and genomic contexts of this radical

trophic shift, which has occurred far less frequently in diatoms than it has in other groups.

Like most other lineages that have lost photosynthesis, apochloritic diatoms maintain

highly reduced plastids and plastid genomes (Kamikawa et al. 2015a) . Their plastids lack

chlorophyll and thylakoids (Kamikawa et al. 2015b) , and their plastid genomes have lost most

photosynthesis-related genes, including all photosystem genes (Kamikawa et al. 2015a) . The

nearly complete set of ATP synthase genes in the plastid genome might function in ATP

hydrolysis, creating a proton gradient that fuels protein import into the plastid (Kamikawa et al.

2015a) . Carbon metabolism is highly compartmentalized in diatoms (Smith et al. 2012) , and the

large number of nuclear-encoded proteins targeted to the plastid point to a highly metabolically

active and, as a result, indispensable (Kamikawa et al. 2017) organelle. Although comparative

genomics is greatly improving our understanding of carbon metabolism in both photosynthetic

(Smith et al. 2012) and nonphotosynthetic diatoms (Kamikawa et al. 2015a, 2017) , the power of

comparative genomics can only be fully leveraged within the framework of an accurate, densely

sampled phylogenetic hypothesis.

We collected, isolated, and cultured several apochloritic Nitzschia species and sequenced

common phylogenetic markers to test whether photosynthesis was lost one or multiple times. A

combined dataset of nuclear, mitochondrial, and plastid genes support monophyly of apochloritic

Nitzschia species, consistent with a single loss of photosynthesis in diatoms. Species were split

between two major subclades that co-occur in habitats with apparently favorable, albeit poorly

defined, conditions. We also sequenced and characterized the plastid genome of Nitzschia sp. and

13

found it to be highly similar to that of another species in the same subclade, indicating rapid

genomic streamlining following loss of photosynthesis in their common ancestor. The results

presented here help frame a number of questions about the evolution, ecology, and genomic

consequences of this rare, radical trophic shift in diatoms.

Materials and Methods

Collection and culturing of apochloritic Nitzschia

All cultures originated from a single composite sample, collected on 10 November 2011

from Whiskey Creek, which is located in Dr. Von D. Mizell-Eula Johnson State Park (formerly

John U. Lloyd State Park), Dania Beach, Florida, USA (26.081330 lat, –80.110783 long). This

site was previously shown to host a diverse assemblage of apochloritic Nitzschia species

(Blackburn et al. 2009a) . Our sample consisted of near-surface plankton collected with a 10 µM

mesh net, submerged sand (1m and 2m depth), and nearshore wet (but unsubmerged) sand. We

selected for nonphotosynthetic species by storing the sample in the dark at room temperature

(21℃) for several days before isolating colorless diatom cells with a Pasteur pipette. Clonal

cultures were grown in the dark at 21℃ on agar plates made with L1+NPM medium (Guillard

1960, Guillard and Hargraves 1993) and 1% Penicillin–Streptomycin–Neomycin solution

(Sigma-Aldrich P4083) to retard bacterial growth.

DNA extraction, PCR, and DNA sequencing

Cells were rinsed with L1 medium and removed from agar plates by pipetting, briefly

centrifuged, then broken with MiniBeadbeater-24 (BioSpec Products). We then extracted DNA

with a Qiagen DNeasy Plant Mini Kit. Nuclear SSU and partial LSU rDNA genes were

14

PCR-amplified and sequenced using published PCR conditions and primer sequences (Alverson

et al. 2007) . PCR and sequencing primers for two mitochondrial markers, cytochrome b ( cob )

and NADH dehydrogenase subunit 1 ( nad1 ), are listed in Table S1. PCRs for mitochondrial

genes used: 1.0–5.0 µL of DNA, 6.5 µL of Failsafe Buffer E (Epicentre Technologies), 0.5 µL of

each primer (20 µM stocks), 0.5 units Taq polymerase, and ddH 2 O to a final volume of 25 µL. In

a few cases, we used a nested PCR strategy to amplify the nad1 gene. PCR conditions for cob

and nad1 genes were as follows: 95 ℃ for 5 minutes, 36 cycles of (95 ℃ for 60 s, 45 ℃ for 60

s, 72 ℃ for 60 s), and a final extension at 72 ℃ for 5 minutes. PCR products were sequenced on

an ABI 3100 capillary sequencer. Raw sequences were assembled and edited with Geneious ver.

7.1.4 (Biomatters Ltd.) and deposited in GenBank (Table 1).

Phylogenetic analyses

In addition to the data generated here, we compiled previously published data for the

plastid 16S and nuclear 28S rDNA genes for Bacillariales to better enable comparisons with

previous studies. We downloaded all Bacillariales sequences from Genbank, checked percent

identity and coverage of each sequence to a local BLAST database comprised of apochloritic

Nitzschia sequences, and reverse-complemented sequences if necessary. We kept Bacillariales

sequences that were at least 20% identical and covered at least 40% of at least one target in the

database (BLAST options: identity_cutoff=20, hsp_coverage_cutoff=40). We kept only the

longest sequence for cases in which multiple downloaded sequences had the same NCBI Taxid

identifier. The total number of sequences meeting these criteria were 46 for the 16S and 116 for

the 28S rDNA genes.

15

We used SSU-ALIGN ver. 0.1 (Nawrocki et al. 2009) to align rDNA sequences (plastid

16S and the d1–d3 region of the nuclear 28S rDNA), using SSU-ALIGN’s built-in covariance

models of secondary structure for bacteria for the 16S alignment and a heterokont-specific

covariance model for the 28S alignment (Nakov et al. 2014) . We used SSU-MASK to remove

poorly aligned regions and used these more conservative alignments for downstream analyses.

The protein-coding cob and nad1 genes from the mitochondrial genome were aligned by hand

with Mesquite ver. 3.31 (Maddison, W. P. and Maddison, D. R. 2008) after color-coding

nucleotide triplets by their conceptual amino acid translations. We built phylogenetic trees from

each individual gene alignment and three concatenated alignments: (1) a concatenation of the

two mitochondrial genes into a cob + nad 1 (heretofore mito dataset), (2) a concatenation of the

two mitochondrial genes and the masked 28S alignment of newly generated 28S sequences

(heretofore mito-lsu dataset), and (3) a concatenation of the two mitochondrial genes and the

masked alignment of all 28S sequences (newly generated and downloaded from GenBank,

heretofore mito-ncbi-lsu dataset). The alignments have been archived in a ZENODO online

repository (https://doi.org/10.5281/zenodo.1211571).

We used RAxML ver. 8.2.4 (Stamatakis 2014) to infer phylogenetic trees from each of

the three concatenated alignments. For each dataset, we performed 10 maximum likelihood

searches to find the best-scoring tree and a rapid bootstrap analysis of 250 pseudoreplicates per

search (Stamatakis et al. 2008) , for a total of 2500 bootstrap samples per alignment. We used the

general time-reversible model (GTR) of nucleotide substitution, and we used a Gamma

distribution to accommodate rate variation across sites within each alignment (GTR+GAMMA,

the default model in RAxML).

16

Plastid genome sequencing and analysis

We sequenced the plastid genome of Nitzschia sp. (Nitz4) using the Illumina HiSeq2000

platform, with a 300-bp library and 90-bp paired-end reads. We removed adapter sequences and

trimmed raw reads with Trimmomatic ver. 0.32 and settings

‘ILLUMINACLIP=<TruSeq_adapters.fasta>:2:30:10, TRAILING=5,

SLIDINGWINDOW=6:18, HEADCROP=9, MINLEN=50’ (Bolger et al. 2014) . We assembled

trimmed reads using Ray ver. 2.3.1 with default settings and k-mer length of 45 (Boisvert et al.

2012) . We assessed the assembly quality with QUAST ver. 2.3 (Gurevich et al. 2013) , confirmed

high genome-wide read coverage by mapping the trimmed reads to the assembly with Bowtie

ver. 0.12.8 (Langmead et al. 2009) , and evaluated these results with SAMtools. We used

DOGMA (Wyman et al. 2004) and ARAGORN (Laslett and Canback 2004) to annotate the

genome. The annotated genome has been archived in GenBank under accession MG273660 . We

used Easyfig (Sullivan et al. 2011) to perform a BLASTN-based synteny comparison of the

plastid genomes of two apochloritic Nitzschia taxa, strains Nitz4 (this study) and NIES-3581

(Kamikawa et al. 2015a) .

Results

Phylogeny of apochloritic Nitzschia

We isolated and cultured four strains (Nitz2, Nitz4, Nitz7, and Nitz8) of apochloritic

Nitzschia from Whiskey Creek (Florida, USA), including both linear and undulate forms (Fig.

1-1). Although cells grew well in culture, we never observed sexual reproduction and, unlike

17

some other apochloritic Nitzschia strains that have been maintained in culture for decades, all of

our strains experienced substantial size reduction and eventually died off.

We sequenced the nuclear 28S rDNA and mitochondrial cob and nad1 genes for 26

Bacillariales taxa, including seven apochloritic strains (Table 1-1). We combined these data with

GenBank sequences to reconstruct phylogenetic trees for individual and concatenated

alignments. The plastid 16S rDNA gene supported monophyly of apochloritic Nitzschia species

(Fig. 1-2A, Bootstrap proportion (BS)=100), though the exceptionally long branch

lengths—driven by decreased GC content—raises the possibility that this grouping simply

reflects shared nucleotide composition (Steel et al. 1993, Galtier and Gouy 1995, Kamikawa et

al. 2015b) . One way to reduce the influence of GC bias in tree inference is to transform the

alignment into purine/pyrimidine (R/Y) coding. By treating all A and G nucleotides as R and all

C and T nucleotides as Y, the GC bias present in some sequences is masked, the base frequencies

are normalized, and the majority of phylogenetic signal originates from transversions (i.e., R ↔

Y). The phylogeny resulting from the R/Y-transformed alignment should better reflect the history

of the lineage rather than the nucleotide composition bias. Performing this transformation for the

plastid 16S alignment, we again found strong support for monophyly of the apochloritic

Nitzschia (bootstrap support = 95%) (Fig. 1-2B). This result suggests that monophyly of

apochloritic Nitzschia species reconstructed with the plastid-encoded 16S gene might not

necessarily be an artifact of the decreased GC content in the plastid genomes of

nonphotosynthetic taxa.

A previous analysis of nuclear 28S d1–d2 rDNA sequences weakly supported

non-monophyly of apochloritic taxa, though monophyly could not be rejected (Kamikawa et al.

18

2015b) . Expanding the analysis to include the larger number of Bacillariales 28S d1–d2

sequences from GenBank and newly sequenced taxa from this study (Table 1-1) returned a

monophyletic grouping of apochloritic Nitzschia (Fig. 1-2C). We also sequenced two

mitochondrial genes, cob and nad1 , and although neither of them individually supported

monophyly of apochloritic taxa (Fig. 1-2 D, E), analysis of a concatenated mitochondrial

cob + nad1 alignment did return a clade of apochloritic Nitzschia (Fig. 1-2F). Concatenated

alignments of nuclear 28S d1–d2 and the two mitochondrial genes supported monophyly of

apochloritic Nitzschia as well, both for an analysis restricted to just newly sequenced taxa (Fig.

1-2G) and one that included both newly sequenced and GenBank taxa (Fig. 2H). In all cases,

branch support for monophyly of apochloritic Nitzschia with nuclear and mitochondrial markers,

analyzed individually or in combination, was less than 70% (Fig. 1-2). The fully labeled trees are

available in Fig. S2.

Plastid genome sequencing and analysis

The plastid genome of Nitzschia sp. Nitz4 maps as a circular chromosome and has the

conserved quadripartite architecture typical of most plastid genomes, with two inverted repeat

(IR) regions and small and large single copy regions (SSC and LSC, respectively). At 67,764 bp

in length, the genome is roughly half the size of plastid genomes from photosynthetic diatoms

(Ruck et al. 2014, Yu et al. 2018) . The genome contains minimal intergenic DNA, with 12 genes

that overlap by anywhere from 1–61 bp in length and another two genes that are immediately

adjacent to one another. (Table S1-2). The genome consists mainly of genes with conserved

housekeeping functions, including 32 tRNA and 6 rRNA genes. All three rRNA and six of the

tRNA genes are present twice in the genome because of their location in the IR region. More

19

than half (45) of the 73 protein-coding genes encode ribosomal proteins or subunits of RNA

polymerase, with ATP synthase genes and ORFs constituting most of the remaining genes. Some

of the ORFs include ones (ycf41, ycf89 and ycf90) that are highly conserved among diatoms

(Ruck et al. 2014, Yu et al. 2018) . The genome also contains subunit of Sec-independent protein

translocator TatC and SecA subunit of Sec-mediated transport system, iron-sulfur clusters

forming proteins SufB and SufC, molecular chaperone dnaK and protease (ClpC). The genome is

highly AT-rich (77.6% A+T).

Similar to the plastid genome of another apochloritic diatom, Nitzschia sp. NIES-3581

(aka IriIs04) (Kamikawa et al. 2015a) , the genome of Nitzschia sp. Nitz4 is missing all genes

encoding subunits of photosystem I and II, proteins of cytochrome b6f complex, carbon fixation

system, porphyrin and thiamine metabolism. As well Nitz4 lacks several conserved ORFs and

membrane translocators subunits, dnaB helicase, chaperonine groEL and ftsH cell division

protein. The genomes of Nitz4 and NIES-3581 are highly similar in size, gene content, gene

order, nucleotide composition, and sequence (Fig. 3). The loss of rps20 in Nitz4, which appears

to have been replaced by a unique ORF (orf122), is among the few minor differences between

the two genomes (Fig. 1-3). Two gene fusions, orf96-atpB and orf122-rpoB, present in Nitz4 are

separated in NIES-3581. Likewise, the overlapping dnaK-tRNA-Arg genes in NIES-3581 are

adjacent but non-overlapping in Nitz4.

20

Discussion

Monophyly of apochloritic diatoms

The small number of species (roughly 20) and narrow taxonomic range (all Bacillariales,

mostly all Nitzschia ) of apochloritic diatoms suggests that these species—the only known

nonphotosynthetic diatoms—are monophyletic, reflecting a single loss of photosynthesis and

transition to obligate heterotrophy in their common ancestor. Previous phylogenetic tests of this

hypothesis were equivocal, however, with the plastid 16S rDNA sequences supporting

monophyly of apochloritic species and the nuclear 28S d1–d2 gene splitting them, albeit with

low bootstrap support, into three separate clades (Kamikawa et al. 2015b) . Further underscoring

the uncertainty in their relationships, the non-normalized, highly AT-rich plastid 16S sequences

may have resulted in long-branch artifacts (Steel et al. 1993, Galtier and Gouy 1995) , and the

nuclear 28S dataset could not reject monophyly of apochloritic species (Kamikawa et al. 2015b) ,

pointing to insufficient signal in this relatively short sequence fragment to address this question.

As efforts to understand the causes and genomic consequences of loss of photosynthesis in

diatoms accelerate (Kamikawa et al. 2015a, 2015b, 2017) , it is important to determine if the shift

occurred one or multiple times. The number, pattern, and timing of shifts has important

implications for understanding of the ease of transition(s) to obligate heterotrophy, the ecological

setting of the transition(s), and, in practical terms, the power of comparative approaches to

resolve these questions—as evolutionary replication, represented in this case by multiple losses

of photosynthesis—maximizes the power of comparative methods to reveal possible mechanisms

underlying these kinds of evolutionary transitions (e.g., Maddison and FitzJohn 2015) . A single

transition would, by contrast, greatly limit the power of comparative methods to uncover the

21

underlying genomic and ecological drivers of the switch away from auto- or mixotrophy to

obligate heterotrophy.

To address this problem, we increased both the number of taxa and genes available for

phylogenetic analyses. We also normalized the plastid 16S rDNA sequences to guard against

artefactual grouping of apochloritic species with highly AT-rich plastid genomes. Although

support was generally low, the best trees inferred from genes representing all three genetic

compartments uniformly supported monophyly of apochloritic Nitzschia (Fig.1- 2). In short, the

best available data support a single loss of photosynthesis in Nitzschia and, by extension, diatoms

as a whole. In light of this, future research efforts can focus attention away from questions that

naturally arise for characters that evolve multiple times (e.g., the role of convergent evolution)

and focus more on firmly placing the apochloritic clade within the broader Bacillariales

phylogeny to help identify, for example, the ecological and mixotrophic characteristics of their

closest relatives. Considerable divergence and phylogenetic structure exists within the

apochloritic clade, with the largest combined-data tree (mito-lsu-new-ncbi) splitting taxa into

two major subclades (Figs. 1-2 and S1-1). The modest level of species diversity, in turn, might

make it feasible to sample each subclade more-or-less exhaustively and apply comparative

approaches that will be useful for understanding the genomic and metabolic consequences of the

switch to heterotrophy, including, for example, the rate of decay of the photosynthetic apparatus.

All of this requires a robust, time-calibrated phylogeny of Bacillariales, which is one of the

largest and most species-rich taxonomic orders of diatoms (Kociolek et al. 2018) . This and other

phylogenetic studies have begun resolving some parts of the Bacillariales tree using small

numbers of commonly used phylogenetic markers (e.g., Lundholm et al. 2002, Rimet et al. 2011,

22

Smida et al. 2014) , but many relationships remain unresolved. With most of the traditional

markers now more-or-less exhausted, effort should turn to much deeper sampling of the nuclear

genome, which appears to hold great promise for resolving relationships across many

phylogenetic scales within diatoms (Parks et al. 2018) .

The ecology and biogeography of apochloritic diatoms

Although relatively few apochloritic Nitzschia species have been formally described and

named, the amount of phylogenetic diversity described in this (Fig. 1-2) and other studies

(Kamikawa et al. 2015b) , suggests that this clade could easily comprise on the order of 20

species. This seems even more probable when considering the relatively modest historical efforts

to collect and characterize apochloritic diatoms. One emerging, and quite striking, theme among

the small number of studies that focused on characterizing species diversity in this group is the

large number and diversity of apochloritic Nitzschia that co-occur over very small spatial scales

(e.g., within a sample from a single site). The diversity is apparent from both morphological and

molecular phylogenetic evidence alike (Figs. 1-1 and 1-2; Blackburn et al. 2009a, Kamikawa et

al. 2015b) . Although some of the apochloritic taxa in our trees came from culture collections

(Table 1-1 and Fig. S1-1), most of the taxa derive from a small number of mangrove-dominated

habitats in Japan (Kamikawa et al. 2015b) and the United States (Figs. 1-1 and 1-2; Blackburn et

al. 2009a) . Both of these sites host multiple (apparently undescribed) apochloritic Nitzschia

species that span the full phylogenetic breadth of the clade (Fig. S1-1). Focused efforts to collect

and culture apochloritic Nitzschia from amenable habitats (e.g., seagrasses and mangroves)

worldwide will show whether this anecdotal trend holds. If it is upheld, the next natural step will

be to determine whether species co-occurrence is made possible by the sheer abundance of local

23

resources in these habitats or rather by resource partitioning among species, either in fine-scale

microhabitats or through specialization on different carbon sources. Of course, both of these

alternatives require knowledge of the organic carbon sources used by these species.

Nevertheless, the phylogeny clearly shows species are not clustered geographically (Fig. S1-1),

which suggests that many or most apochloritic diatom species have broad, presumably

worldwide, geographic distributions and that the local species pool at any one site is the product

of dispersal, not in situ speciation.

Plastid genome reduction in apochloritic diatoms

Although probably different species, the two Nitzschia strains with sequenced plastid

genomes belong to the same subclade (Fig. S1A, bottom) and so, in the context of the entire

apochloritic clade, are very close relatives. Their close relationship was also evident in their

plastid genomes, which were highly similar in nearly all respects. Additional sampling across the

entire apochloritic clade will show whether all species share the same fundamental, highly

reduced plastid genome—indicative of rapid genomic streamlining following the loss of

photosynthesis in their common ancestor—or whether decay of the plastid genome is ongoing in

some taxa. Similar to some green algae (Knauf and Hachtel 2002) , diatoms have retained a

nearly full set of ATP synthase genes whose products, instead of functioning in ATP hydrolysis,

presumably generate a proton gradient for tat-dependent protein translocation across the

thylakoid membrane (Kamikawa et al. 2015a) . This model underscores both the

compartmentalized nature of carbon metabolism in diatoms (Smith et al. 2012, Kamikawa et al.

2017) and, consequently, the indispensable nonphotosynthetic plastids in diatoms.

24

In addition to understanding the specific consequences for the plastid genome to

following the loss of photosynthesis, a fuller understanding of these plastid genomes can shed

light on photosynthetic diatom plastids as well. For example, the retention several conserved

ORFs in these nonphotosynthetic genomes—in the context of near wholesale loss of the

photosynthetic apparatus—strongly suggests these conserved, diatom-specific ORFs have

functions that are unrelated, or only peripherally related, to photosynthesis. These ORFs include

ycf41, ycf89, and ycf90. Finally, the loss of photosynthesis can have cascading effects on rates of

sequence evolution all three genomes (Nickrent et al. 1998) . Branch lengths based on plastid,

mitochondrial, and nuclear genes showed that the AT-driven rate acceleration in the plastid

genomes of apochloritic species appears to be restricted to the plastid genome alone (Fig.1- 2).

This does not rule out that there have not been other effects on the mitochondrial and nuclear

genomes including, for example, gene content and whether some of the missing plastid genes

have been transferred to the nucleus.

Conclusions

The discovery that apochloritic Nitzschia are monophyletic represents an important step

forward in understanding the loss of photosynthesis and switch to obligate heterotrophy in

diatoms—a transition that has occurred just once in a lineage of some 100,000 species of

photoautotrophs. A clearer view of this relationship highlights new research avenues and

priorities, including more intensive taxon sampling within Bacillariales to identify the closest

relatives of the apochloritic clade, which almost certainly evolved from a mixotrophic ancestor

(Hellebust and Lewin 1977, Figueroa-Martinez et al. 2015) . The exact carbon sources, modes of

25

carbon uptake and utilization, and the degree of carbon specialization within and across species

remains unclear for photosynthetic and nonphotosynethetic Nitzschia alike. Bacillariales features

a diverse set of taxa with fascinating biology, motivating the development of excellent genomic

resources for this group (Basu et al. 2017, Kamikawa et al. 2017, Mock et al. 2017) . A nuclear

genome sequence for a nonphotosynthetic Nitzschia could help address several of these

outstanding questions, leading to hypotheses that can be directly tested in a group of diatoms that

has proven highly amenable to experimental manipulation (e.g., Lewin and Lewin 1967, Azam

and Volcani 1974, McGinnis and Sommerfeld 2000) .

26

References

Alverson, A.J., Jansen, R.K. & Theriot, E.C. 2007. Bridging the Rubicon: Phylogenetic analysis reveals repeated colonizations of marine and fresh waters by thalassiosiroid diatoms. Mol. Phylogenet. Evol. 45:193–210.

Archibald, J.M. 2009. The puzzle of plastid evolution. Curr. Biol. 19:R81–8. Azam, F. & Volcani, B.E. 1974. Role of silicon in diatom metabolism. Arch. Microbiol. 101:1–8.

Barkman, T.J., McNeal, J.R., Lim, S.H., Coat, G., Croom, H.B., Young, N.D. & DePamphilis, C.W. 2007. Mitochondrial DNA suggests at least 11 origins of parasitism in angiosperms and reveals genomic chimerism in parasitic plants. BMC Evol. Biol. 7:248.

Basu, S., Patil, S., Mapleson, D., Russo, M.T., Vitale, L., Fevola, C., Maumus, F. et al. 2017. Finding a partner in the ocean: Molecular and evolutionary bases of the response to sexual cues in a planktonic diatom. New Phytol. 215:140–56.

Blackburn, M.V., Hannah, F. & Rogerson, A. 2009a. First account of apochlorotic diatoms from mangrove waters in Florida. J. Eukaryot. Microbiol. 56:194–200.

Blackburn, M.V., Hannah, F. & Rogerson, A. 2009b. First account of apochlorotic diatoms from intertidal sand of a south Florida beach. Estuar. Coast. Shelf Sci. 84:519–26.

Blouin, N.A. & Lane, C.E. 2012. Red algal parasites: Models for a life history evolution that leaves photosynthesis behind again and again. Bioessays . 34:226–35.

Boisvert, S., Raymond, F., Godzaridis, E., Laviolette, F. & Corbeil, J. 2012. Ray Meta: Scalable de novo metagenome assembly and profiling. Genome Biol. 13:R122.

Bolger, A.M., Lohse, M. & Usadel, B. 2014. Trimmomatic: A flexible trimmer for Illumina sequence data. Bioinformatics . 30:2114–20.

Donaher, N., Tanifuji, G., Onodera, N.T., Malfatti, S.A., Chain, P.S.G., Hara, Y. & Archibald, J.M. 2010. The complete plastid genome sequence of the secondarily nonphotosynthetic alga Cryptomonas paramecium : Reduction, compaction, and accelerated evolutionary rate. Genome Biol. Evol. 1:439–48.

Field, C.B., Behrenfeld, M.J., Randerson, J.T. & Falkowski, P. 1998. Primary production of the biosphere: Integrating terrestrial and oceanic components. Science . 281:237–40.

Figueroa-Martinez, F., Nedelcu, A.M., Smith, D.R. & Reyes-Prieto, A. 2015. When the lights go out: The evolutionary fate of free-living colorless green algae. New Phytol. 206:972–82.

Galtier, N. & Gouy, M. 1995. Inferring phylogenies from DNA sequences of unequal base compositions. Proc. Natl. Acad. Sci. U. S. A. 92:11317–21.

27

Goff, L.J., Moon, D.A., Nyvall, P., Stache, B., Mangin, K. & Zuccarello, G. 1996. The evolution of parasitism in the red algae: Molecular comparisons of adelphoparasites and their hosts. J. Phycol. 32:297–312.

Guillard, R.R.L. 1960. A mutant of Chlamydomonas moewusii lacking contractile vacuoles. J. Eukaryot. Microbiol. 7:262–8.

Guillard, R.R.L. & Hargraves, P.E. 1993. Stichochrysis immobilis is a diatom, not a chrysophyte. Phycologia . 32:234–6.

Gurevich, A., Saveliev, V., Vyahhi, N. & Tesler, G. 2013. QUAST: Quality assessment tool for genome assemblies. Bioinformatics . 29:1072–5.

Hadariová, L., Vesteg, M., Hampl, V. & Krajčovič, J. 2017. Reductive evolution of chloroplasts in non-photosynthetic plants, algae and protists. Curr. Genet.

Hellebust, J.A. & Lewin, J. 1977. Heterotrophic nutrition. In Werner, D. [Ed.] The Biology of Diatoms . University of California Press, pp. 169–97.

Kamikawa, R., Moog, D., Zauner, S., Tanifuji, G., Ishida, K.-I., Miyashita, H., Mayama, S. et al. 2017. A non-photosynthetic diatom reveals early steps of reductive evolution in plastids. Mol. Biol. Evol. 34:2355–66.

Kamikawa, R., Tanifuji, G., Ishikawa, S.A., Ishii, K.I., Matsuno, Y., Onodera, N.T., Ishida, K.I. et al. 2015a. Proposal of a twin arginine translocator system-mediated constraint against loss of ATP synthase genes from nonphotosynthetic plastid genomes. Mol. Biol. Evol. 32:2598–604.

Kamikawa, R., Yubuki, N., Yoshida, M., Taira, M., Nakamura, N., Ishida, K.I., Leander, B.S. et al. 2015b. Multiple losses of photosynthesis in Nitzschia (Bacillariophyceae). Phycological Res. 63:19–28.

Keeling, P.J. 2010. The endosymbiotic origin, diversification and fate of plastids. Philos. Trans. R. Soc. Lond. B Biol. Sci. 365:729–48.

Knauf, U. & Hachtel, W. 2002. The genes encoding subunits of ATP synthase are conserved in the reduced plastid genome of the heterotrophic alga Prototheca wickerhamii . Mol. Genet. Genomics . 267:492–7.

Kociolek, J.P., Balasubramanian, K., Blanco, S., Coste, M., Ector, L., Liu, Y., Kulikovskiy, M. et al. 2018. DiatomBase. Available at: http://www.diatombase.org/ (last accessed April 5, 2018).

Kurihara, A., Abe, T., Tani, M. & Sherwood, A.R. 2010. Molecular phylogeny and evolution of red algal parasites: a case study of Benzaitenia , Janczewskia , and Ululania (Ceramiales). J. Phycol. 46:580–90.

28

Langmead, B., Trapnell, C., Pop, M. & Salzberg, S.L. 2009. Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol. 10:R25.

Laslett, D. & Canback, B. 2004. ARAGORN, a program to detect tRNA genes and tmRNA genes in nucleotide sequences. Nucleic Acids Res. 32:11–6.

Lewin, J. & Lewin, R.A. 1967. Culture and nutrition of some apochlorotic diatoms of the genus Nitzschia . Microbiology . 46:361–7.

Lundholm, N., Daugbjerg, N. & Moestrup, Ø. 2002. Phylogeny of the Bacillariaceae with emphasis on the genus Pseudo-nitzschia (Bacillariophyceae) based on partial LSU rDNA. Eur. J. Phycol. 37:115–34.

Maddison, W. P. and Maddison, D. R. 2008. Mesquite: A modular system for evolutionary analysis. Evolution . 62:1103–18.

Maddison, W.P. & FitzJohn, R.G. 2015. The unsolved challenge to phylogenetic correlation tests for categorical characters. Syst. Biol. 64:127–36.

Marin, B. 2004. Origin and fate of chloroplasts in the euglenoida. Protist . 155:13–4.

Martin-Cereceda, M., Roberts, E.C., Wootton, E.C., Bonaccorso, E., Dyal, P., Guinea, A., Rogers, D. et al. 2010. Morphology, ultrastructure, and small subunit rDNA phylogeny of the marine heterotrophic flagellate Goniomonas aff. amphinema . J. Eukaryot. Microbiol. 57:159–70.

McFadden, G.I., Reith, M.E., Munholland, J. & Lang-Unnasch, N. 1996. Plastid in human parasites. Nature . 381:482.

McGinnis, K.M. & Sommerfeld, M.R. 2000. PLASTID FATTY ACID BIOSYNTHESIS IN THE DIATOMS NITZSCHIA ALBA AND NITZSCHIA LAEVIS. J. Phycol. 36:46–46.

Mock, T., Otillar, R.P., Strauss, J., McMullan, M., Paajanen, P., Schmutz, J., Salamov, A. et al. 2017. Evolutionary genomics of the cold-adapted diatom Fragilariopsis cylindrus . Nature . 541:536–40.

Nakov, T., Ruck, E.C., Galachyants, Y., Spaulding, S.A. & Theriot, E.C. 2014. Molecular phylogeny of the Cymbellales (Bacillariophyceae, Heterokontophyta) with a comparison of models for accommodating rate variation across sites. Phycologia . 53:359–73.

Nawrocki, E.P., Kolbe, D.L. & Eddy, S.R. 2009. Infernal 1.0: Inference of RNA alignments. Bioinformatics . 25:1335–7.

29

Nickrent, D.L., Duff, R.J., Colwell, a. E., Wolfe, a. D., Young, N.D., Steiner, K.E. & DePamphilis, C.W. 1998. Molecular phylogenetic and evolutionary studies of parasitic plants. In Molecular Systematics of Plants II. DNA Sequencing. Springer US, Boston, MA, p. 211–41.

Parks, M.B., Wickett, N.J. & Alverson, A.J. 2018. Signal, Uncertainty, and Conflict in Phylogenomic Data for a Diverse Lineage of Microbial Eukaryotes (Diatoms, Bacillariophyta). Mol. Biol. Evol. 35:80–93.

Pringsheim, E.G. 1956. Micro-organisms from decaying seaweed. Nature . 178:480–1.

Raven, J.A. 1997. Phagotrophy in phototrophs. Limnol. Oceanogr. 42:198–205.

Reyes-Prieto, A., Moustafa, A. & Bhattacharya, D. 2008. Multiple genes of apparent algal origin suggest ciliates may once have been photosynthetic. Curr. Biol. 18:956–62.

Rimet, F., Kermarrec, L., Bouchez, A., Hoffmann, L., Ector, L. & Medlin, L.K. 2011. Molecular phylogeny of the family Bacillariaceae based on 18S rDNA sequences: focus on freshwater Nitzschia of the section Lanceolatae. Diatom Res. 26:273–91.

Ruck, E.C., Nakov, T., Jansen, R.K., Theriot, E.C. & Alverson, A.J. 2014. Serial gene losses and foreign DNA underlie size and sequence variation in the plastid genomes of diatoms. Genome Biol. Evol. 6:644–54.

Rumpf, R., Vernon, D., Schreiber, D. & Birky, C.W. 1996. Evolutionary consequences of the loss of photosynthesis in Chlamydomonadaceae: Phylogenetic analysis of Rrn18 (18S rDNA) in 13 Polytoma strains (Chlorophyta). J. Phycol. 32:119–26.

Saldarriaga, J.F., Taylor, F.J.R., Keeling, P.J. & Cavalier-Smith, T. 2001. Dinoflagellate nuclear SSU rRNA phylogeny suggests multiple plastid losses and replacements. J. Mol. Evol. 53:204–13.

Smida, D.B., Lundholm, N., Kooistra, W.H.C.F., Sahraoui, I., Ruggiero, M.V., Kotaki, Y., Ellegaard, M. et al. 2014. Morphology and molecular phylogeny of Nitzschia bizertensis sp. nov.—A new domoic acid-producer. Harmful Algae . 32:49–63.

Smith, S.R., Abbriano, R.M. & Hildebrand, M. 2012. Comparative analysis of diatom genomes reveals substantial differences in the organization of carbon partitioning pathways. Algal Research . 1:2–16.

Stamatakis, A. 2014. RAxML version 8: A tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics . 30:1312–3.

Stamatakis, A., Hoover, P. & Rougemont, J. 2008. A rapid bootstrap algorithm for the RAxML Web servers. Syst. Biol. 57:758–71.

30

Steel, M.A., Lockhart, P.J. & Penny, D. 1993. Confidence in evolutionary trees from biological sequence data. Nature . 364:440–2.

Sullivan, M.J., Petty, N.K. & Beatson, S.A. 2011. Easyfig: A genome comparison visualizer. Bioinformatics . 27:1009.

Tartar, A. & Boucias, D.G. 2004. The non-photosynthetic, pathogenic green alga Helicosporidium sp. has retained a modified, functional plastid genome. FEMS Microbiol. Lett. 233:153–7.

Těšitel, J. 2016. Functional biology of parasitic plants: A review. Plant Ecol. Evol. 149:5–20.

Tuchman, N.C., Schollett, M.A., Rier, S.T. & Geddes, P. 2006. Differential heterotrophic utilization of organic compounds by diatoms and bacteria under light and dark conditions. Hydrobiologia . 561:167–77.

Tyler, B.M., Tripathy, S., Zhang, X., Dehal, P., Jiang, R.H.Y., Aerts, A., Arredondo, F.D. et al. 2006. Phytophthora genome sequences uncover evolutionary origins and mechanisms of pathogenesis. Science . 313:1261–6.

Waller, R.F. & McFadden, G.I. 2005. The apicoplast: A review of the derived plastid of apicomplexan parasites. Curr. Issues Mol. Biol. 7:57–79.

Wyman, S.K., Jansen, R.K. & Boore, J.L. 2004. Automatic annotation of organellar genomes with DOGMA. Bioinformatics . 20:3252–5.

Yan, D., Wang, Y., Murakami, T., Shen, Y., Gong, J., Jiang, H., Smith, D.R. et al. 2015. Auxenochlorella protothecoides and Prototheca wickerhamii plastid genome sequences give insight into the origins of non-photosynthetic algae. Sci. Rep. 5:14465.

Yu, M., Ashworth, M.P., Hajrah, N.H., Khiyami, M.A., Sabir, M.J., Alhebshi, A.M., Al-Malki, A.L. et al. 2018. Evolution of the plastid genomes in diatoms. Adv. Bot. Res.

31

Table 1-1. Taxa and sources of DNA sequences analyzed in this study.

Genus Species Culture ID Apoch- lorotic cob nad1 lsu 16s

Nitzschia sp. ccmp2144 MH017326 MH017300 MH017352 Nitzschia dippelii AJA014-6 MH017327 MH017301 MH017353

Hantzschia spectabilis UTEX FD269 MH017328 MH017302 MH017354

Hantzschia elongata UTEX FD421 MH017329 MH017303 MH017355

Hantzschia amphioxys AJA013-11 MH017330 MH017304 MH017356

Fragilariopsis cylindrus ccmp1022 JGI proj ID: 16035

JGI proj ID: 16036

Pseudo- nitzschia sp. ccmp1447 MH017331 MH017305 MH017357 Nitzschia frustulum ccmp1532 MH017332 MH017306 MH017358 Nitzschia cf. ovalis ccmp1118 MH017333 MH017307 MH017359 Cylindrotheca closterium ccmp1855 MG271845 MG271845 Psammodictyon constrictum ccmp576 MH017334 MH017308 MH017360 Nitzschia laevis ccmp1092 MH017335 MH017309 MH017361 Nitzschia sp. OHI12.10 MH017336 MH017310 MH017362 Nitzschia cf. frequens ccmp1500 MH017337 MH017311 MH017363 Nitzschia pusilla ccmp2526 MH017338 MH017312 MH017364 Nitzschia cf. pusilla ccmp581 MH017339 MH017313 MH017365 Nitzschia sp. ccmp2533 MH017340 MH017314 MH017366 Nitzschia draveiliensis AJA010-17 MH017341 MH017315 MH017367

Nitzschia palea var. debilis AJA013-2 MH017342 MH017316 MH017368

Nitzschia gracilis AJA010-53 MH017343 MH017317 MH017369 Nitzschia capitellata AJA012-22 MH017344 MH017318 MH017370 Nitzschia alba ccmp2426 yes MH017345 MH017319 MH017371 Nitzschia leucosigma ccmp2197 yes MH017346 MH017320 MH017372 MH017379 Nitzschia sp. ccmp579 yes MH017347 MH017321 MH017373 Nitzschia sp. Nitz2 yes MH017348 MH017322 MH017377 Nitzschia sp. Nitz4 yes MH017349 MH017323 MH017375 MH017378 Nitzschia sp. Nitz7 yes MH017350 MH017324 MH017376 Nitzschia sp. Nitz8 yes MH017351 MH017325 MH017374 MH017380

32

Figure 1-1. Light micrographs of newly sequenced nonphotosynthetic Nitzschia species. Clockwise from top: Nitzschia sp. nitz2, Nitzschia sp. nitz7, Nitzschia sp. nitz8, Nitzschia sp. nitz4.

33

Figure 1-2. Phylogenetic trees inferred from plastid 16S (A), plastid 16S transformed into purine/pyrimidine (R/Y) coding (B), nuclear 28S d1–d2 genes for all newly sequenced taxa and

34

data from GenBank [”lsu (new + ncbi)”, C], mitochondrial cob (D), mitochondrial nad1 (E), concatenated cob and nad1 genes [”mito (cob + nad1 )”, F], concatenated cob, nad1, and nuclear 28S d1–d2 genes for all newly sequenced taxa [”mito (cob + nad1 ) + lsu (new)”, G], and a concatenated alignment of cob, nad1, and nuclear 28S d1–d2 genes for taxa from this study and GenBank [”mito (cob + nad1 ) + lsu (new + ncbi)”, H]. For the phylogenies in panels C and H, we removed branches shorter than 0.00001 units for clarity. The full phylogenies are available in Supplementary Figure 1. Thicker black branches correspond to apochlorotic taxa. White points identify nodes with bootstrap support >70%.

35

Figure 1-3. Conserved synteny, gene content, and sequence in the plastid genomes of two nonphotosynthetic diatoms in the genus Nitzschia, Nitz4 (this study) and NIES-3581 (Kamikawa et al. 2015a). Unlabeled genes are shared between the two species.

36

Table S1-1. Primers used to amplify and sequence cob and nad1 fragments for study taxa.

Primer Name Primer Sequence (5’-3’)

CobnF GGAGTTTYGGYTCTTTAGCWGG

CobnR GGHARAAAATACCACTCWGGVAC

Nad1nF a ATGGGAGCAATYCAAAGRCG

Nad1nR a CATTTCCAACCTAAATACATTAATTG

Nad1F-mod b AAAGACGACGAGGWCCWAATGTKATAGGTT

Nad1R-mod b CATTAATTGGTCATAYCGGTAWCGWGG

a Forward and reverse primers for initial amplification reaction. b Forward and reverse primers for second amplification reaction when nested PCR was performed.

37

Table S1-2. Overlapping and adjacent genes in the plastid genomes of Nitzschia sp. nitz4 and Nitzschia sp. NIES-3581. Groups of overlapping or adjacent genes are separated by empty rows with the coordinates showing the degree of overlap. All but two groups are shared between the two genomes. Nitzschia sp. Nitz4 Nitzschia sp. NIES-3581

Gen name

Gene coordinates

Overlapping

Adjacent Gene name

Gene coordinates

Overlapping

Adjacent

rpl19 Complement (3679..3993) rpl19

Complement (3681..3995)

orf96 Complement (3986..4276) + urf97

Complement (3970..4260) +

atpB Complement (4267..5694) +

orf122 6041..6409

rpoB 6406..9513 +

rpoC1 9517..11580 rpoC1 9410..11479

rpoC2 11577..14900 + rpoC2 11476..14820 +

rps2 14938..15627 rps2 14858..15547

sufB 15624..17072 + sufB 15544..16992 +

sufC 17072..17815 + sufC 16992..17735 +

atpG 18859..19311 atpG 18777..19229

atpF 19298..19732 + atpF 19216..19650 +

atpA 20300..21805 atpA 20206..21720

rpl35 21783..21977 + rpl35 21714..21908 +

rps12 Complement (36097..36471) rps12

Complement (36024..36398)

rpl31 Complement (36471..36680) + rpl31

Complement (36398..36604) +

38

Table S1-2. (Cont.) Nitzschia sp. Nitz4 Nitzschia sp. NIES-3581

Gen name

Gene coordinates

Overlapping

Adjacent Gene name

Gene coordinates

Overlapping

Adjacent

rps9

Complement (36687..37091)

rps9

Complement (36612..37016)

rpl13

Complement (37088..37537)

+

rpl13

Complement (37013..37462) +

rpoA Complement (37534..38466) + rpoA

Complement (37459..38391) +

rps11 Complement (38487..38891) rps11

Complement (38412..38822)

rps13 Complement (38830..39273) + rps13

Complement (38755..39198) +

secY Complement (39409..40455) secY

Complement (39334..40398)

rps5 Complement (40427..40930) + rps5

Complement (40370..40873) +

rps17 Complement (43317..43556) rps17

Complement (43288..43527)

rpl29 Complement (43546..43782) + rpl29

Complement (43517..43756) +

dnaK 48083..49930

tRNA- Arg 49930..50003 +

TOTAL: 12 2 TOTAL: 10 3

39

Fig S1-1. (Cont.)

40

Fig S1-1. (Cont.)

41

Fig S1-1. (Cont.)

42

Fig S1-1. (Cont.)

43

Fig S1-1. (Cont.)

44

Fig S1-1. (Cont.)

45

Fig S1-1. (Cont.)

46

Figure S1-1. Phylogenetic trees for plastid 16S (A), plastid 16S transformed into purine/pyrimidine (R/Y) coding (B), mitochondrial cob (C), nad1 (D), a densely sampled 28S d1–d2 matrix with sequences from this study and GenBank (E), cob and nad1 combined (F), and a large combined nuclear and mitochondrial gene matrix with sequences from this study and GenBank (G). In the 16S tree, Nitzschia sp. NIES-3581 is identified by its synonym, iriis04.

47

Chapter 2. Core carbon metabolism and characterization of a β-ketoadipate pathway

inferred from the genome of a non-photosynthetic diatom (Bacillariophyta)

Abstract

Although most of the tens of thousands of diatom species are photoautotrophs, many

mixotrophic species can also use extracellular sources of organic carbon, and one small lineage

of obligate heterotrophs has lost the ability to photosynthesize and requires extracellular organic

carbon for growth. We lack an understanding of the exact sources of carbon used by these

species, however. We sequenced the genome of a non-photosynthetic diatom, Nitzschia sp., and

used it to develop a comprehensive model of carbon metabolism in these species. The genome is

relatively small (31 Mbp), gene dense, and contains a β-ketoadipate pathway, a pathway known

mostly from bacteria and fungi. The β-ketoadipate pathway potentially allows diatoms to

metabolize lignin-derived aromatic compounds, the products of which are predicted to be

delivered to mitochondria and fed directly into the Krebs cycle. Genes in this pathway are also

present in genomes of some photosynthetic diatoms, suggesting that this mode of carbon

utilization is an ancestral feature of diatoms, allowing them to exploit derivatives of one of the

most abundant sources of organic carbon on the planet.

48

Introduction

Diatoms are microscopic unicellular algae that are widely distributed throughout marine

and fresh waters. They are prolific photosynthesizers responsible for some 20% of global

primary production (Field et al. 1998, Smetacek 1999) . Diatom plastids trace back to a secondary

endosymbiosis between a bi-flagellated host—the common ancestor of the chromalveolate

clade—and a red algal endosymbiont (Archibald 2009) . As a result, diatom plastids are

surrounded by four membranes, including two relic primary plastid membranes, which are

remnants of red algal plasmalleme, and the diatom endoplasmic membrane (Stoebe and Maier

2002) . The first sequenced diatom genome captured this history, revealing many genes

transferred from the red algal endosymbiont into the host nuclear genome (Armbrust et al. 2004) .

Most functions in the diatom plastids are, in fact, carried out by proteins that are encoded in the

nuclear genome and synthesized in the cytoplasm. To be properly targeted and transported into

the plastid, plastid-targeted proteins must contain N-terminal bipartite presequences with signal

and transit peptide domains separated by a cleavage site that contains highly conserved amino

acid motif (Lang et al. 1998, Gruber et al. 2007) .

Carbon metabolism is highly compartmentalized in diatoms. Diatom plastids possess a

full Calvin cycle, but lack complete pentose phosphate pathway (PPP), whereas, and the lower

part of the glycolytic pathway takes place in the mitochondrion. Many genes associated with

primary metabolism enzymes exist in multiple copies in the genome and differ in their targeting

sequences (Kroth et al. 2008, Smith et al. 2012) . Models of carbon metabolism in diatoms are

based on photosynthetic species, but a broad diversity of diatoms, mostly in the pennate lineage,

are able to use external sources of carbon as well, probably as a means of maintaining their

49

growth in low light conditions (Tuchman et al. 2006) . One small subclade of raphid pennate

diatoms in the genus Nitzschia have lost the ability to photosynthesize altogether and exist now

as free-living heterotrophs. The 20 or so apochloritic diatoms trace back to a single loss of

photosynthesis in their common ancestor (Onyshchenko et al. 2018) . These species are often

found in habitats rich in organic matter, including mangrove forests (Blackburn et al. 2009,

Kamikawa et al. 2015, Onyshchenko et al. 2018) and decaying seaweeds (Pringsheim 1956,

Lewin and Lewin 1967) , potentially providing some clues about the sources of organic carbon

available to them.

We sequenced, assembled, and annotated the genome of one non-photosynthetic diatom,

Nitzschia sp. (Nitz4) to understand how these species acquire and metabolize organic carbon for

energy. Apochloritic diatoms also maintain colorless plastids with highly reduced plastid

genomes (Kamikawa et al. 2016, Onyshchenko et al. 2018) . A model of carbon metabolism in

these plastids suggests that this organelle maintains a high degree of carbon metabolic activity

and biochemical complexity even in the absence of a photosynthetic apparatus (Kamikawa et al.

2017) . The goal of our study was to use the Nitzschia sp. Nitz4 genome to identify possible

sources of exogenous carbon and to develop a comprehensive, whole-cell model of carbon

metabolism in nonphotosynthetic diatoms.

Materials and Methods

Collection and culturing of Nitzschia sp.

We collected a composite sample on 10 November 2011 from Whiskey Creek, which is

located in Dr. Von D. Mizell-Eula Johnson State Park (formerly John U. Lloyd State Park), Dania

50

Beach, Florida, USA (26.081330 latitude, –80.110783 longitude). The sample consisted of

near-surface plankton collected with a 10 µM mesh net, submerged sand (1 m and 2 m depth),

and nearshore wet (but unsubmerged) sand. We selected for non-photosynthetic species by

storing the sample in the dark at room temperature (21℃) for several days before isolating

colorless diatom cells with a Pasteur pipette. Clonal cultures were grown in the dark at 21℃ on

agar plates made with L1+NPM medium (Guillard 1960, Guillard and Hargraves 1993) and 1%

Penicillin–Streptomycin–Neomycin solution (Sigma-Aldrich P4083) to retard bacterial growth.

DNA and RNA extraction and sequencing

We rinsed cells with L1 medium and removed them from agar plates by pipetting, lightly

centrifuged them, and then disrupted them with a MiniBeadbeater-24 (BioSpec Products). We

extracted DNA with a Qiagen DNeasy Plant Mini Kit. We sequenced total extracted DNA from a

single culture strain, Nitzschia sp. Nitz4, with the Illumina HiSeq2000 platform housed at the

Beijing Genomics Institute, with a 500-bp library and 90-bp paired-end reads. We separately

extracted total RNA with a Qiagen RNeasy kit, and sequenced a 300-bp Illumina TruSeq RNA

library using the the Illumina HiSeq2000 platform.

Genome and transcriptome assembly and annotation

A total of 15.4 GB of pair-end DNA reads were recovered and used to assemble the Nitz4

nuclear, plastid, and mitochondrial genomes. We used FastQC (ver. 0.11.5) (Andrews 2010) to

check read quality then used ACE (Sheikhizadeh and de Ridder 2015) to correct predicted

sequencing errors. We subsequently trimmed and filtered the reads using Trimmomatic (ver.

0.32) (Bolger et al. 2014) with settings “ILLUMINACLIP=<TruSeq_adapters.fa>:2:40:15,

LEADING=2 TRAILING=2, SLIDINGWINDOW=4:2, MINLEN=30, TOPHRED64.” We

51

assembled the genome using RAY (ver. 2.3.1) (Boisvert et al. 2012) with a range of k-mer

lengths (15, 21, 27, 33, 39, 45, 51, 59, 65) to determine which k-mer length maximized N50,

maximum scaffold length, and total assembly length. Based on these criteria, the assembly based

on a k-mer length of 45 was judged to be best (Fig. S2-1). We then used bowtie2 (ver. 2.2.8)

(Langmead and Salzberg 2012) remove reads that mapped to the previously assembled plastid

genome (Onyshchenko et al. 2018) . The remaining nuclear and mitochondrial reads were

reassembled following the strategy outlined above with a k-mer length of 45. To identify

mitochondrial contigs, we used NCBI-BLASTN to search all of the assembled scaffolds against

a database of diatom mitochondrial genes. This search returned a single mitochondrial scaffold.

We then mapped and removed all mitochondrial reads as described above for the plastid genome.

The remaining set of organelle-filtered reads were assembled again with RAY and range of

k-mer lengths, and the assembly that used a k-mer length of 45 was again the best one (Fig.

S2-2). We used Blobtools (ver. 1) (Laetsch and Blaxter 2017) to visualize the assembly and

verify that the final nuclear assembly was free of organelle and contaminating bacterial contigs.

We then used AGOUTI (ver 0.3.3) (Zhang et al. 2016) to scaffold together contigs that could be

joined based on paired-end information in the RNA-seq reads. The AGOUTI step used the fourth

Maker-based genome annotation (see “Annotation” methods below).

We used REAPR (Hunt et al. 2013) with the smalt map aligner and default parameter

settings to check the quality of our de novo nuclear genome assembly. REAPR uses

read-mapping data to assess read and read-pair depth, as well as read-pair conflicts, across the

genome, breaking potentially misassembled contigs that have, for example, local concentrations

of read-pair conflicts or coverage values that fall outside of an expected distribution. We found

52

that original, unbroken assembly did not differ critically from the REAPR-produced assembly, so

we proceeded with original assembly.

RNA extraction, sequencing, read processing, and assembly of RNA-seq data followed

Parks and Nakov et al. (2017) .

Genome annotation

We used the Maker pipeline (ver. 2.31.8) (Cantarel et al. 2008) to annotate the Nitz4

nuclear genome. In order to successively improve the annotation quality, we ran Maker a total of

six times with default settings unless stated otherwise. We used the assembled Nitz4

transcriptome (Maker’s expressed sequence tag [EST] evidence) and the predicted proteome of

Fragilariopsis cylindrus (GCA_001750085.1) (Maker’s protein homology evidence) to inform

the annotation. We used RepeatMasker (ver. open-4.0.7) (Chen 2004) to generate a repeat library

for Nitz4, which was also used by Maker for the annotation. We used Augustus (ver. 3.2.2) for de

novo gene prediction with settings “max_dna_len = 200,000” and “min_contig = 300.” We

trained Augustus using the set of annotated genes from Phaeodactylum tricornutum (ver. 2),

which we filtered as follows: (1) remove “hypothetical” and “predicted” proteins, (2) remove all

but one splice variant of a gene, (3) remove genes with no introns, and (4) remove genes that

overlap with neighboring gene models or the 1000 bp flanking regions of adjacent genes, as

these regions are included in the training set along with the enclosed gene. If possible, we

generated UTR annotations for the selected gene models by subtracting 5’ and 3’CDS

coordinates from the corresponding 5’ and 3’ end coordinates of the associated mRNA sequence;

among these, we only retained those with UTRs that extended ≥25 bp beyond both the 5’ and 3’

ends of the CDS. The final filtered training set included a total of 726 genes.

53

We generated a separate set of gene models for UTR training by using all of the filtering

criteria outlined above except that we retained intronless genes, as we assumed intron presence

or absence was less relevant for training the UTR annotation parameters. In addition, we retained

only those genes with UTRs that were ≥40 bp in length. In total, our UTR training set included

531genes. We trained and optimized the Augustus gene prediction parameters (with no UTRs) on

the first set of 726 genes and optimized UTR prediction parameters with the second set of 531

genes. We then used these parameters to perform a series of six successive gene predictions

within Maker. We assessed the first Maker annotation for completeness using BUSCO (ver. 2.0)

(Simão et al. 2015) with the protist database (Table 2-1) . Following recommendations of the

developers, for the five subsequent Maker runs, we used a different ab initio gene predictor,

SNAP (Korf 2004) , trained with Maker-generated gene models from the previous run. Although

the fifth SNAP-based Maker run discovered more genes, the number of complete BUSCOs

decreased by one, so we discarded this assembly and used the previous one (the fifth Maker run,

the fourth SNAP-based run) as the final assembly.

Scripts used for genome assembly, annotation and MAKER gene training set search are

available at https://github.com/Nastassiia/Nitz4_annot_methods_clean.git .

Construction of orthologous clusters

We used Orthofinder (ver. 1.1.4) (Emms and Kelly 2015) with default parameters to build

orthologous clusters from the complete set of predicted proteins from the genomes of Nitz4,

Fragilariopsis cylindrus (GCA_001750085.1), Phaeodactylum tricornutum

( GCA_000150955.2 ), Cyclotella nana ( GCA_000149405.2 ), and the transcriptome of another

Nitzschia species (Nitz2144) (accession forthcoming). We tried to characterize species-specific

54

Nitz4 genes by using NCBI-BLASTP to search orthogroups that consisted exclusively of Nitz4

genes against NCBI’s nonredundant (nr) database (GenBank release 223.0 or 224.0).

Characterization of carbon metabolism genes

We performed a thorough manual annotation of primary carbon metabolism genes in the

Nitz4 genome. We started with a set of core carbon metabolic genes in diatoms (e.g., Smith et al.

2012, Kamikawa et al. 2017) , downloaded all genes from GenBank’s nr database (release 223.0

or 224.0) using the annotation as the search term, and filtered them to include just RefSeq

accessions. As we characterized carbon metabolic pathways, we expanded our searches as

necessary to include genes that were predicted to be involved in those pathways. We constructed

local BLAST databases of Nitz4 genome scaffolds, predicted proteins, and assembled transcripts

and used the NCBI-BLAST (ver. 2.6.0) package to separately search the set of sequences for a

given carbon metabolism gene against the three Nitz4 databases. In most cases, each gene had

thousands of annotated sequences on GenBank, so if the GenBank sequences for a given gene

were, in fact, homologous and accurately annotated, then we expected Nitz4 homologs to

likewise return thousands of strong hits, and this was often the case. We did not consider further

any putative carbon genes with relatively few query hits (e.g., tens of weak hits vs. the thousands

of strong hits in verified, correctly annotated sequences). We extracted the Nitz4 subject matches

and used NCBI-BLASTX or BLASTP to search the transcript or predicted protein sequence

against NCBI’s nr protein database to verify the annotation. Nitz4 subject matches in scaffold

regions were checked for overlap with annotated genes, and if the subject match did overlap with

an annotated gene, the largest annotated CDS in this region was extracted and BLASTed to the nr

55

protein database to verify the annotation. As necessary, we repeated this procedure for other

diatom species used to construct clusters of orthologous genes.

Prediction of protein localization

We used the software programs SignalP (Petersen et al. 2011) , ASAFind (Gruber et al.

2015) , ChloroP (Emanuelsson et al. 1999) , TargetP (Emanuelsson et al. 2007) , MitoProt (Claros

1995) , and HECTAR (Gschloessl et al. 2008) to predict whether proteins were targeted to the

plastid, mitochondrion, cytoplasm, or endoplasmic reticulum (ER) for the secretory pathway. Our

approach was slightly modified from that of Traller et al. (2016) . Proteins with a SignalP-,

ASAFind, and HECTAR-predicted plastid signal peptide were classified as plastid targeted. We

placed less weight on ChloroP target predictions, which are optimized for targeting to plastids

bound by two membranes and not the four-membrane plastids found in diatoms. Proteins

predicted as mitochondrial-targeted by any two of the TargetP, MitoProt, and HECTAR programs

were classified as localized to the mitochondrion. Those proteins in the remaining set were

classified as ER targeted if they had ER signal peptides predicted by SignalP and HECTAR.

Most of the remaining proteins were cytoplasmic.

Results and Discussion

Genome characteristics

A total of 74,996,078 100-bp paired-end DNA reads assembled into 4,447 nuclear

scaffolds totaling 30.4 Mbp in length and with an average read depth of 180x. AGOUTI joined

350 contigs that were broken within transcribed regions, reducing the number of scaffolds to

4,097 and increasing the assembly size to 30.7 Mbp in length. Half of the Nitz4 genome is

56

contained within 195 scaffolds, each one longer than 43,578 bp (Table 2-2, N50). The average

scaffold length is 7,507 bp, and the single largest scaffold is 340,060 bp in length. The Nitz4

genome is similar in size to other small diatom genomes (Table 1), but it is the smallest so far

sequenced from Bacillariales, the lineage that also includes F. cylindrus (61 Mbp) and

Pseudo-nitzschia multistriata (55 Mbp) (Table 2-2). The genomic GC content is 47.8%, which is

similar to most other diatoms (Table 2-2).

The fifth iteration of Maker resulted in 9,235 gene models, the vast majority (9,017) of

which were supported by protein or EST evidence. The gene models included a total of 209 of

234 protist BUSCOs (Table 2-1). Although Nitz4 has fewer genes than all other diatoms

sequenced to date, it also has the largest BUSCO count, suggesting that the smaller overall

number of genes is not due to incomplete sequencing of the genome (Table 2-1). Genes models

from C. nana , F. cylindrus , Nitzschia sp. 2144, and Nitz4 clustered into 10,834 multi-sequence (≥

2 gene models) orthogroups. A total of 8,304 of the Nitz4 genes were assigned to orthogroups,

and 21 genes fell into six Nitz4-specific orthogroups. The remaining genes did not cluster with

any other sequences, either from Nitz4 or another species.

Central carbon metabolism

Glycolysis, gluconeogenesis, and pyruvate metabolism —Glycolysis is a core cytosolic pathway

for degradation of glucose to pyruvate, which can then be targeted to mitochondria for ATP

production and conversion into a range of other compounds that can, in turn, be used for further

energy production. Through gluconeogenesis, several pyruvate metabolizing enzymes convert

pyruvate back into hexoses when the energy demands of the cell have been met, or when

necessary, into glycolysis carbon intermediates for anabolism. In diatoms, including Nitz4, the

57

latter half of the glycolytic pathway can occur in both the cytosol and mitochondria, and part of

the pathway can be carried out in the plastid (Kroth et al. 2008, Smith et al. 2012) . Overall, Nitz4

contains fewer copies of genes encoding glycolytic enzymes compared to other diatoms. These

include the plastid-targeted FBP, FBI, TPI genes and cytosolic GADPH and PGK genes (Fig.

S2-4). The first enzyme involved in glycolysis, glucokinase (GK), exists as a single copy in the

Nitz4 genome and has no targeting signal, indicating that the first step of glycolysis likely occurs

in the cytosol. Similar to other diatoms, we did not find a hexokinase gene in Nitz4. The second

glycolytic enzyme, glucose-6-phosphate isomerase ( GPI), was predicted to be localized to both

the plastid and the cytosol, though none of the three phosphofructokinase-1 (PFK) genes

possessed plastid targeting signals. Enzymes for subsequent steps in glycolysis,

fructose-bisphosphate aldolase (FBA) and pyruvate kinase (PK), are localized to both the

cytosolic and plastid compartments, as predicted in other diatoms as well (Kroth et al. 2008,

Smith et al. 2012, Traller et al. 2016) .

As in other diatoms (Smith et al. 2012) , enzymes catalyzing reactions in the latter half of

glycolysis (from triosephosphate isomerase [TPI] to pyruvate kinase [ PK]) were also predicted to

be localized to mitochondria (Fig. S2-3), which appears to be a conserved feature across diatoms

(Kroth et al. 2008, Smith et al. 2012) . The Nitz4 genome contains two copies of

phosphoglycerate kinase (PGK), which catalyzes the seventh reversible step of glycolysis and

has a possible role in gluconeogenesis. Surprisingly, however, neither copy contains a

mitochondrial targeting signal—instead, one copy is predicted to be plastid-targeted and the

other is cytosolic. We verified the absence of a mitochondrial PGK thorough careful manual

searching of both the scaffolds and the transcriptome. Lack of a mitochondrial PGK suggests that

58

Nitz4 either lacks half of the mitochondrial glycolysis/gluconeogenesis pathway or that one or

both of the plastid and cytosolic PGK genes are, in fact, dual-targeted to the mitochondrion. In

the true, and seemingly unlikely absence, of a mitochondrial-targeted PGK, the mitochondrion

would be missing a single critical enzyme for glycolysis and gluconeogenesis. We did not find

mitochondrial phosphoenolpyruvate (PEP) exporters, which would suggest that this intermediate

compound is shuttled into a cytosolic gluconeogenic pathway after conversion from pyruvate.

We then characterized the presence and localization of genes in the so-called “pyruvate

hub” (Smith et al. 2012) , which describes the set of genes involved in conversion of pyruvate for

reuse in gluconeogenesis or other metabolic pathways (Fig. S2-3). Overall, the “pyruvate hub” in

Nitz4 is reduced by comparison to photosynthetic diatoms both in the number of different

enzymes present in the genome and in protein localization, which is restricted mostly to the

mitochondrial compartment in Nitz4. Among the possible enzymes involved in pyruvate

conversion to phosphoenolpyruvate ( PEP), Nitz4 possesses only the pyruvate carboxylase ( PC)

and PEPCK enzymes, which convert pyruvate through an intermediate compound, oxaloacetate

( OAA). Both enzymes, which are encoded by single copy genes, were predicted to be localized

to the mitochondrion. Another unidirectional gluconeogenesis enzyme, fructose bisphosphatase

(FBP), is targeted to the cytosol and plastids, as predicted for other diatoms (Kroth et al. 2008,

Smith et al. 2012) . Nitz4 is missing glucose-6-phosphatase (G6Pase), which is responsible for

the third committed gluconeogenic reaction and, as a result, Nitz4 cannot produce free glucose in

this pathway.

Like other diatoms (Kroth et al. 2008, Smith et al. 2012) , Nitz4 targets enzymes involved

in pyruvate conversion to the mitochondrion. As previously stated, the PC and PEPCK enzymes

59

that convert pyruvate to PEP through OAA are restricted to mitochondrial compartment, but the

reverse reaction—conversion of PEP back to OAA by phosphoenolpyruvate carboxylase

(PEPC)—is carried out in the cytosol. This differs from other diatoms, which localize pyruvate

carboxylase (PC) to the plastid and PEPC to the plastid/periplastidial space (Smith et al. 2012) .

Phosphoenolpyruvate synthetase (PEPS) and phosphate dikinase (PPDK), which can catalyze the

first reaction of gluconeogenesis, are not present in Nitz4 (Fig. S2-3). Malic enzyme (ME),

which is involved in pyruvate metabolism and C4-photosynthesis in photosynthetic diatoms, is

missing from Nitz4. The role of ME in diatom metabolism is unclear (Smith et al. 2012) , the the

apparent loss of ME in Nitz4 could be a direct consequence of the loss of photosynthesis or a

secondary effect related to other metabolic changes. It is also hard to predict how the inability to

convert malate directly to pyruvate might influence carbonic flux through the TCA cycle. Malate

dehydrogenase enzyme (MDH) is targeted to the mitochondrion, which is expected considering

its indispensable role in the TCA cycle. MDH might also function in gluconeogenesis by

converting malate to OAA and, subsequently, to pyruvate in conjunction with PEPCK. MDH is

targeted to the mitochondrion in Nitz4, ruling out any involvement in the Asp–Malate shuttle

and, consequently, the transfer of reducing equivalents (Mikulášová 1998) .

Apart from PFK, a complete set of glycolytic enzymes are targeted to the plastid in Nitz4,

a feature that is conserved in other diatoms (e.g., C. cryptica and F. cylindrus ) as well. In the

absence of a photosynthesis-derived carbon source in the plastid, Nitz4 may import glycolysis

intermediates into the plastid for further metabolism.

Detailed annotations of genes constituting Nitzschia sp. Nitz4 central metabolic pathways

are available at Appendix 1.

60

A β-ketoadipate pathway in diatoms

Experimental data have shown that mixotrophic diatoms apparently can use a range of

exogenous carbon sources (Hellebust and Lewin 1977, Tuchman et al. 2006) . The exact source

of carbon in strictly heterotrophic, nonphotosynthetic diatoms is unclear, however, and

identifying the carbon species used by these diatoms was a principal goal of this study. The

genome sequences allows us to compile the gene models and make such predictions based on the

presence of specific carbon-related genes and pathways or, alternatively, based on the expansions

of gene families related to specific modes of carbon metabolism relative to photosynthetic

species. Species in the pennate diatom clade (to which Nitz4 belongs) are mixotrophic, and

mixotrophy is naturally thought to be the intermediate trophic strategy between autotrophy and

heterotrophy (Figueroa-Martinez et al. 2015) , so genome comparisons to other raphid pennate

diatoms may conceal obvious features related to obligate vs. facultative heterotrophy. For

example, major differences between mixotrophs and strict heterotrophs may involve other kinds

of changes in, for example, transcription levels of or post-translational modifications.

Orthogroup clustering revealed six Nitz4-specific orthogroups with multiple gene

models, and we could not identify or ascribe functions to most of these groups. A systematic

survey of orthogroups enriched for Nitz4 homologs (≥ 3 Nitz4 sequences) returned one

species-specific cluster (OG0006126) with five Nitz4 genes that had strong matches to genes

encoding protocatechuate dioxygenase (PCA), an enzyme involved in the degradation of

aromatic compounds, specifically through intradiol ring cleavage of a hydroxylated derivative of

benzoate, protocatechuic acid. Four of these genes are located on relatively long scaffolds with

identifiable diatoms genes (Figure 2-1). A further search of the genomic scaffolds revealed

61

several additional copies of unannotated PCA enzymes, and BLASTP searches to GenBank’s nr

protein database returned to hits to protocatechuate 3 4-dioxygenase (P3,4O) enzymes. With no

prior knowledge of the presence or role of this enzyme in diatoms, we focused a great deal of

attention characterizing this gene, its role in the larger protocatechuate pathway, and the overall

contribution of this pathway to central carbon metabolism in Nitz4.

The major role of PCA is degradation of aromatic substrates through the β-ketoadipate

pathway (Stanier and Ornston 1973, Harwood and Parales 1996, Fuchs et al. 2011) , which is

known mostly from soil fungi and eubacteria. The β-ketoadipate does not function in

protocatechuate degradation per se , rather it serves as a funneling pathway for further

metabolism of a variety of aromatic compounds commonly found in nature (Fuchs et al. 2011) .

Aromatic rings are a common feature of many organic compounds, and organisms degrade these

compounds into a range of aromatic intermediates (e.g., protocatechuate, catechol, gentisate, and

benzoyl-CoA), which are then subject to ring-cleavage reactions and further degradation through

a limited number of metabolic pathways. Thus, derivatives of aromatic environmental pollutants

and natural substrates, such as lignin, released by plants may be subject to catabolism through

protocatechuate intermediates (Harwood and Parales 1996, Masai et al. 2007, Fuchs et al. 2011,

Wells and Ragauskas 2012) . Lignin is of particular interest here because it comprises a large

fraction of plant biomass and is, as a result, one of the most abundant biopolymers in nature.

Lignin is an unordered heterogeneous polymer consisting of aromatic rings and is highly

resistant to biodegradation, requiring a suite of microbial enzymes and non-enzymatic steps for

its decomposition (Janusz et al. 2017) . Several lignin derived monomers, which are the products

of decaying plants, are converted to protocatechuate and metabolized through one branch of the

62

β-ketoadipate pathway (Stanier and Ornston 1973, Harwood and Parales 1996) . In summary,

lignin from degrading plant material represents a major source of naturally occurring aromatic

compounds that could feed directly into aromatic degradation pathways, including the

β-ketoadipate pathway.

In addition to PCA, we discovered that Nitz4 has a nearly complete set of enzymes for

the β-ketoadipate pathway, including single copies of 3-carboxymuconate lactonizing enzyme

(CMLE), 4-carboxymuconolactone decarboxylase (CMD), and beta-ketoadipate enol-lactone

hydrolase (ELH) (Table 2-3, Fig. 2-2) . PCA is the first enzyme in the protocatechuic branch of

the β-ketoadipate pathway (Fig. 2-2) and is responsible for ortho-cleavage of the activated

benzene ring between two hydroxyl groups, which results in the production of

β-carboxymuconic acid (Fig. 2-2). CMLE, CMD, and ELH catalyze subsequent steps of the

pathway after fission of the benzene ring (Fig. 2-2). The last two enzymes in the pathway,

3-ketoadipate:succinyl-CoA transferase (TR) and 3-ketoadipyl-CoA thiolase (TH), convert

β-ketoadipate to TCA intermediates, succinyl-CoA and acetyl-CoA, are missing from the Nitz4

genome. However, these two enzymes are highly similar to ones involved in central carbon

metabolism—Succinyl-CoA:3-ketoacid CoA transferase (SCOT) and 3-ketoacyl-CoA thiolase

(3-KAT), respectively (Parales and Harwood n.d., Kaschabek et al. 2002) . Thus, we predict that

these two core carbon metabolism enzymes substitute in for the last two steps of the

β-ketoadipate pathway . This hypothesis will require experimental validation, but the presence of

intact genes in the upper portion of the pathway strongly suggests than an entire pathway exists,

and these enzymes are the best candidates for fulfilling these roles. Moreover, both TH and TR

are strongly predicted to be targeted to the mitochondrion.

63

Products of the two last steps of β-ketoadipate pathway, succinyl-CoA and acetyl-CoA,

are TCA cycle intermediates. The predicted mitochondrial localization of these steps therefore

strongly suggests that these products are funneled directly into the TCA cycle, providing carbon

skeletons for generation of ATP. This hypothesis requires, however, transport of β-ketoadipate

into the mitochondrion, a process that may involve the mitochondrial 2-oxoglutarate/malate

antiporter (Zeman et al. 2016) , which is usually a component of the malate/aspartate shuttling

system. Both cytoplasmic and mitochondrial MDH enzymes are necessary for this shuttling

system, however, and most diatoms, including Nitz4, only have a mitochondrial MDH. This

suggests that the 2-oxoglutarate/malate transporter in Nitz4 may function β-ketoadipate

transport into the mitochondrion. In addition, the diatom also needs to import protocatechuate

into the cell. All diatom genomes, including Nitz4, contain a putative 4-hydroxybenzoate

transporter, which can reportedly transport protocatechuate (3,4-dihydroxybenzoate) (Nichols

and Harwood n.d.) . Alternatively, protocatechuate might be synthesized from other metabolites

transported into the cell, but we found no evidence for enzymes involved in degradation of

protocatechuate precursors. In summary, it is difficult to predict which diatom transport system is

involved in protocatechuate acquisition or whether protocatechuic acid is, in fact, the specific

aromatic compound that is transported into the diatom cell. Experimental evidence will be

necessary to test these hypotheses.

We found evidence for all or part of the β-ketoadipate pathway in all sequenced diatom

genomes as well as the transcriptome of Nitz2144. Fragilariopsis cylindrus lacks both TR and

3-ketoacid:succinyl-CoA transferase, whereas P. tricornutum and C. nana lack PCA, the first

enzyme in the pathway. Phaeodactylum tricornutum is also missing ELH (Fig. 2-2). Overall,

64

only the two Nitzschia species, Nitz4 and Nitz2144, possessed a complete set of enzymes for the

β-ketoadipate pathway. Given the presence of β-ketoadipate genes across diatoms, we wanted to

determine whether Nitz4 was enriched for one or more of the genes in this pathway. The first

gene in this pathway that we discovered, PCA, catalyzes aromatic ring fission, which is the first

and potentially most difficult step in the β-ketoadipate pathway because of the high resonance

energy and, therefore, enhanced chemical stability of the carbon ring structure in protocatechuic

acid (Fuchs et al. 2011) . Consequently, this first critical reaction is one potential “bottleneck” in

the β-ketoadipate pathway. Interestingly, Nitz4 possess fully 10 copies of the PCA gene, whereas

other diatom genomes contain four or fewer copies. Major expansion of the PCA gene family

may suggest that Nitz4 is more efficient than photosynthetic diatom at overcoming the bottleneck

reaction of the β-ketoadipate pathway.

Conclusions

The goal of our study was to comprehensively assess core carbon metabolism in a

non-photosynthetic diatom, one of a few such species in a lineage of some 100,000

photoautotrophic species (Mann and Vanormelingen 2013, Onyshchenko et al. 2018) . The

genome sequence revealed the presence of a β-ketoadipate pathway that appears to have been

completed through co-option of core carbon metabolism proteins. The genomic hypotheses

raised here require follow-up validation with metabolomic experiments. The β-ketoadipate is

mostly known from fungi and bacteria (Harwood and Parales 1996) , so its presence in

Nitz4—and possibly across diatoms as a whole—raises important questions about the origin and

ancestral function of this pathway in diatoms. CMD is absent in fungi, which instead possess

65

3-carboxymuconolactone hydrolase (CMH) converting carboxymuconolactone (Harwood and

Parales 1996) . The presence of CMD, rather than CMH, in Nitz4, together with relatively strong

sequence similarity of Nitz4 CMD to bacterial CMDs suggests that the β-ketoadipate pathway in

diatoms may be of bacterial rather than fungal origin. A full assessment of the origins of this

pathway in diatoms will require detailed phylogenetic analyses of each of the enzymes in the

pathway.

The presence of this pathway raises the possibility that mixotrophic diatoms, and

obligately heterotrophic diatoms like Nitz4, are able to use lignin-derived aromatic compounds

for growth, allowing them to capitalize on one of the most abundant sources of organic carbon

worldwide. Experimental data will be necessary to test and validate this hypothesis, including,

for example determination of whether diatoms can import protocatechuic acid and whether its

derivatives are shuttled directly into the TCA cycle, allowing us to draw a direct line between

plant-derived environmental carbon and ATP production in nonphotosynthetic diatoms.

66

References

Andrews, Simon. 2010. “FastQC: A Quality Control Tool for High Throughput Sequence Data.”

Archibald, J. M. 2009. “The Puzzle of Plastid Evolution.” Current Biology: CB 19 (2): R81–88.

Armbrust, E. Virginia, John A. Berges, Chris Bowler, Beverley R. Green, Diego Martinez, Nicholas H. Putnam, Shiguo Zhou, et al. 2004. “The Genome of the Diatom Thalassiosira Pseudonana : Ecology, Evolution, and Metabolism.” Science 306 (5693): 79–86.

Blackburn, Michele V., Fiona Hannah, and Andrew Rogerson. 2009. “First Account of Apochlorotic Diatoms from Mangrove Waters in Florida.” The Journal of Eukaryotic Microbiology 56 (2): 194–200.

Boisvert, Sébastien, Frédéric Raymond, Elénie Godzaridis, François Laviolette, and Jacques Corbeil. 2012. “Ray Meta: Scalable de Novo Metagenome Assembly and Profiling.” Genome Biology 13 (12): R122.

Bolger, Anthony M., Marc Lohse, and Bjoern Usadel. 2014. “Trimmomatic: A Flexible Trimmer for Illumina Sequence Data.” Bioinformatics 30 (15): 2114–20.

Cantarel, Brandi L., Ian Korf, Sofia M. C. Robb, Genis Parra, Eric Ross, Barry Moore, Carson Holt, Alejandro Sánchez Alvarado, and Mark Yandell. 2008. “MAKER: An Easy-to-Use Annotation Pipeline Designed for Emerging Model Organism Genomes.” Genome Research 18 (1): 188–96.

Chen, Nansheng. 2004. “Using RepeatMasker to Identify Repetitive Elements in Genomic Sequences.” Current Protocols in Bioinformatics / Editoral Board, Andreas D. Baxevanis ... [et Al.] Chapter 4 (May): Unit 4.10.

Claros, M. G. 1995. “MitoProt, a Macintosh Application for Studying Mitochondrial Proteins.” Computer Applications in the Biosciences: CABIOS 11 (4): 441–47.

Emanuelsson, Olof, Søren Brunak, Gunnar von Heijne, and Henrik Nielsen. 2007. “Locating Proteins in the Cell Using TargetP, SignalP and Related Tools.” Nature Protocols 2 (4): 953–71.

Emanuelsson, O., H. Nielsen, and G. von Heijne. 1999. “ChloroP, a Neural Network-Based Method for Predicting Chloroplast Transit Peptides and Their Cleavage Sites.” Protein Science: A Publication of the Protein Society 8 (5): 978–84.

Emms, David M., and Steven Kelly. 2015. “OrthoFinder: Solving Fundamental Biases in Whole Genome Comparisons Dramatically Improves Orthogroup Inference Accuracy.” Genome Biology 16 (August): 157.

67

Field, C. B., M. J. Behrenfeld, J. T. Randerson, and P. Falkowski. 1998. “Primary Production of the Biosphere: Integrating Terrestrial and Oceanic Components.” Science 281 (5374): 237–40.

Figueroa-Martinez, Francisco, Aurora M. Nedelcu, David R. Smith, and Reyes-Prieto Adrian. 2015. “When the Lights Go out: The Evolutionary Fate of Free-Living Colorless Green Algae.” The New Phytologist 206 (3): 972–82.

Fuchs, Georg, Matthias Boll, and Johann Heider. 2011. “Microbial Degradation of Aromatic Compounds – from One Strategy to Four.” Nature Reviews. Microbiology 9 (11): 803–16.

Gruber, Ansgar, Gabrielle Rocap, Peter G. Kroth, E. Virginia Armbrust, and Thomas Mock. 2015. “Plastid Proteome Prediction for Diatoms and Other Algae with Secondary Plastids of the Red Lineage.” The Plant Journal: For Cell and Molecular Biology 81 (3): 519–28.

Gruber, Ansgar, Sascha Vugrinec, Franziska Hempel, Sven B. Gould, Uwe-G Maier, and Peter G. Kroth. 2007. “Protein Targeting into Complex Diatom Plastids: Functional Characterisation of a Specific Targeting Motif.” Plant Molecular Biology 64 (5): 519–30.

Gschloessl, Bernhard, Yann Guermeur, and J. Mark Cock. 2008. “HECTAR: A Method to Predict Subcellular Targeting in Heterokonts.” BMC Bioinformatics 9 (September): 393.

Guillard, Robert R. L. 1960. “A Mutant of Chlamydomonas Moewusii Lacking Contractile Vacuoles.” The Journal of Eukaryotic Microbiology 7 (3). Wiley Online Library: 262–68.

Guillard, R. R. L., and P. E. Hargraves. 1993. “ Stichochrysis Immobilis Is a Diatom, Not a Chrysophyte.” Phycologia 32 (3). The International Phycological Society: 234–36.

Harwood, C. S., and R. E. Parales. 1996. “The Beta-Ketoadipate Pathway and the Biology of Self-Identity.” Annual Review of Microbiology 50: 553–90.

Hellebust, Johan A., and Joyce Lewin. 1977. “Heterotrophic Nutrition.” In The Biology of Diatoms , edited by Dietrich Werner, 13:169–97. University of California Press.

Hunt, Martin, Taisei Kikuchi, Mandy Sanders, Chris Newbold, Matthew Berriman, and Thomas D. Otto. 2013. “REAPR: A Universal Tool for Genome Assembly Evaluation.” Genome Biology 14 (5): R47.

Janusz, Grzegorz, Anna Pawlik, Justyna Sulej, Urszula Swiderska-Burek, Anna Jarosz-Wilkolazka, and Andrzej Paszczynski. 2017. “Lignin Degradation: Microorganisms, Enzymes Involved, Genomes Analysis and Evolution.” FEMS Microbiology Reviews 41 (6): 941–62.

68

Kamikawa, Ryoma, Daniel Moog, Stefan Zauner, Goro Tanifuji, Ken-Ichiro Ishida, Hideaki Miyashita, Shigeki Mayama, et al. 2017. “A Non-Photosynthetic Diatom Reveals Early Steps of Reductive Evolution in Plastids.” Molecular Biology and Evolution 34 (9). academic.oup.com: 2355–66.

Kamikawa, Ryoma, Goro Tanifuji, Sohta A. Ishikawa, Ken-Ichiro Ishii, Yusei Matsuno, Naoko T. Onodera, Ken-Ichiro Ishida, et al. 2016. “Proposal of a Twin Aarginine Translocator System-Mediated Constraint against Loss of ATP Synthase Genes from Nonphotosynthetic Plastid Genomes.” Molecular Biology and Evolution 33 (1). academic.oup.com: 303.

Kamikawa, Ryoma, Naoji Yubuki, Masaki Yoshida, Misaka Taira, Noriaki Nakamura, Ken-Ichiro Ishida, Brian S. Leander, et al. 2015. “Multiple Losses of Photosynthesis in Nitzschia (Bacillariophyceae).” Phycological Research 63 (1): 19–28.

Kaschabek, Stefan R., Bernd Kuhn, Dagmar Müller, Eberhard Schmidt, and Walter Reineke. 2002. “Degradation of Aromatics and Chloroaromatics by Pseudomonas Sp. Strain B13: Purification and Characterization of 3-Oxoadipate:succinyl-Coenzyme A (CoA) Transferase and 3-Oxoadipyl-CoA Thiolase.” Journal of Bacteriology 184 (1): 207–15.

Korf, Ian. 2004. “Gene Finding in Novel Genomes.” BMC Bioinformatics 5 (May): 59.

Kroth, Peter G., Anthony Chiovitti, Ansgar Gruber, Veronique Martin-Jezequel, Thomas Mock, Micaela Schnitzler Parker, Michele S. Stanley, et al. 2008a. “A Model for Carbohydrate Metabolism in the Diatom Phaeodactylum Tricornutum Deduced from Comparative Whole Genome Analysis.” PloS One 3 (1): e1426.

Laetsch, Dominik R., and Mark L. Blaxter. 2017. “BlobTools: Interrogation of Genome Assemblies.” F1000Research 6 (July). https://doi.org/ 10.12688/f1000research.12232.1 .

Lang, Markus, Kirk E. Apt, and Peter G. Kroth. 1998. “Protein Transport into ‘Complex’ Diatom Plastids Utilizes Two Different Targeting Signals.” The Journal of Biological Chemistry 273 (47): 30973–78.

Langmead, Ben, and Steven L. Salzberg. 2012. “Fast Gapped-Read Alignment with Bowtie 2.” Nature Methods 9 (4): 357–59.

Lewin, Joyce, and R. A. Lewin. 1967. “Culture and Nutrition of Some Apochlorotic Diatoms of the Genus Nitzschia.” Microbiology 46 (3). Microbiology Society: 361–67.

Mann, David G., and Pieter Vanormelingen. 2013. “An Inordinate Fondness? The Number, Distributions, and Origins of Diatom Species.” The Journal of Eukaryotic Microbiology 60 (4): 414–20.

Masai, Eiji, Yoshihiro Katayama, and Masao Fukuda. 2007. “Genetic and Biochemical Investigations on Bacterial Catabolic Pathways for Lignin-Derived Aromatic Compounds.” Bioscience, Biotechnology, and Biochemistry 71 (1): 1–15.

69

Mikulášová, D. 1998. “Malate Dehydrogenase: Distribution, Function and Properties.” Gen . Physiol . Biophy S 17: 193–121.

Nichols, Nancy N., and Caroline S. Harwood. n.d. “PcaK, a High-Affinity Permease for the Aromatic Compounds 4-Hydroxybenzoate and Protocatechuate from Pseudomonas Putida .”

Onyshchenko, Anastasiia, Elizabeth C. Ruck, Teofil Nakov, and Andrew J. Alverson. 2018. “A Single Loss of of Photosynthesis in Diatoms.” bioRxiv . https://doi.org/ 10.1101/298810 . In Review .

Parales, Rebecca E., and Caroline S. Harwood. n.d. “Succinyl-Coenzyme Transferase in Pseudomonas Putida .”

Parks, Matthew B., Teofil Nakov, Elizabeth C. Ruck, Norman J. Wickett, and Andrew J. Alverson. 2017. “Phylogenomics Reveals an Extensive History of Genome Duplication in Diatoms (Bacillariophyta).” bioRxiv . https://doi.org/ 10.1101/181115 .

Petersen, Thomas Nordahl, Søren Brunak, Gunnar von Heijne, and Henrik Nielsen. 2011. “SignalP 4.0: Discriminating Signal Peptides from Transmembrane Regions.” Nature Methods 8 (10): 785–86.

Pringsheim, E. G. 1956. “Micro-Organisms from Decaying Seaweed.” Nature 178 (4531): 480–81.

Sheikhizadeh, Siavash, and Dick de Ridder. 2015. “ACE: Accurate Correction of Errors Using K-Mer Tries.” Bioinformatics 31 (19): 3216–18.

Simão, Felipe A., Robert M. Waterhouse, Panagiotis Ioannidis, Evgenia V. Kriventseva, and Evgeny M. Zdobnov. 2015. “BUSCO: Assessing Genome Assembly and Annotation Completeness with Single-Copy Orthologs.” Bioinformatics 31 (19): 3210–12.

Smetacek, V. 1999. “Diatoms and the Ocean Carbon Cycle.” Protist 150 (1): 25–32.

Smith, Sarah R., Raffaela M. Abbriano, and Mark Hildebrand. 2012. “Comparative Analysis of Diatom Genomes Reveals Substantial Differences in the Organization of Carbon Partitioning Pathways.” Algal Research 1 (1): 2–16.

Stanier, R. Y., and L. N. Ornston. 1973. “The Beta-Ketoadipate Pathway.” Advances in Microbial Physiology 9 (0): 89–151.

Stoebe, Bettina, and Uwe-G Maier. 2002. “One, Two, Three: Nature’s Tool Box for Building Plastids.” Protoplasma 219 (3-4): 123–30.

70

Traller, Jesse C., Shawn J. Cokus, David A. Lopez, Olga Gaidarenko, Sarah R. Smith, John P. McCrow, Sean D. Gallaher, et al. 2016a. “Genome and Methylome of the Oleaginous Diatom Cyclotella Cryptica Reveal Genetic Flexibility toward a High Lipid Phenotype.” Biotechnology for Biofuels 9 (November): 258.

Tuchman, Nancy C., Marc A. Schollett, Steven T. Rier, and Pamela Geddes. 2006. “Differential Heterotrophic Utilization of Organic Compounds by Diatoms and Bacteria under Light and Dark Conditions.” Hydrobiologia 561 (1). Kluwer Academic Publishers: 167–77.

Wells, Tyrone, Jr, and Arthur J. Ragauskas. 2012. “Biotechnological Opportunities with the β-Ketoadipate Pathway.” Trends in Biotechnology 30 (12): 627–37.

Zeman, Igor, Martina Neboháčová, Gabriela Gérecová, Kornélia Katonová, Eva Jánošíková, Michaela Jakúbková, Ivana Centárová, et al. 2016. “Mitochondrial Carriers Link the Catabolism of Hydroxyaromatic Compounds to the Central Metabolism in Candida Parapsilosis .” G3 6 (12): 4047–58.

Zhang, Simo V., Luting Zhuo, and Matthew W. Hahn. 2016. “AGOUTI: Improving Genome Assembly and Annotation Using Transcriptome Data.” GigaScience 5 (1): 31.

71

Table 2-1. Nitzschia sp. Nitz4 genome annotation progress statistics

Annotation № Gene models number

Complete BUSCOs

Complete and Duplicated BUSCOs

1 8676 202 5

2 (SNAP HMM1) 9291 208 5

3 (SNAP HMM2) 9324 208 5

4 (SNAP HMM3) 9327 208 5

5 (after agouti rescaffolding; SNAP HMM4) 9235 209 5

FINAL annotation

6 (after agouti rescaffolding; SNAP HMM5) 9295 208 5

Table 2-2. General genome annotation statistics for analyzed diatom species Genome feature Cyclotella

nana Cyclotella cryptica

Phaeodacty- lum tricornutum

Fragilariopsis cylindrus

Nitzschia sp. Nitz4

Pseudonitzschia multistriata

Genome size (Mbp) 32.4 161.7 27.4 61.1 30.7 55.1

GC content (%) 47 43 48.8 38.8 47.8 46.4

Protein-coding genes 11,673 21,121 10,408 18,111 9,235 not annotated

Complete/Duplicated BUSCOs

202/196 198/37 201/192 193/7 209/5 not annotated

Average gene size (bp)

992 1471 1621 1575 1953 not annotated

Gene density (genes per Mbp)

360 131 380 296 301 not annotated

Reference Armbrust et al. 2004

Traller et al. 2016

Bowler et al. 2008

Mock et al. 2017

This study GCA_900005105.1

72

Table 2-3. Diatom β-ketoadipate pathway annotation for analyzed species Species name Gene ID/ Scaffold name Approximate

gene location coordinates

Orthogroup ID

PCA (protocatechuate 3,4-dioxygenase)

Nitzschia sp. Nitz4

augustus_masked-agouti_scaf_108-processed-gene-0.1-mRNA-1 OG0006126

augustus_masked-scaffold-1189-processed-gene-0.1-mRNA-1 OG0006126

augustus_masked-scaffold-393-processed-gene-0.2-mRNA-1 OG0006126

augustus_masked-scaffold-393-processed-gene-0.9-mRNA-1 OG0006126

maker-scaffold-393-augustus-gene-0.15-mRNA-1 OG0006126

agouti_scaf_71 65156-66154 -

agouti_scaf_82 25240-26280 -

scaffold-320 103557-104515 -

scaffold-357 23928-24971 -

scaffold-420 11602-12576 -

F. cylindrus

OEU10566.1 OG0003380

OEU05863.1 OG0003380

OEU16075.1 OG0003380

KV784367.1_FRACYscaffold_15 963661-965124 -

T. pseudonana

XP_002290083.1 -

P. tricornutum

0 0 -

Nitzschia sp. Nitz2144

TRINITY_DN15091_c0_g1_i1|m.24867_nitz2144 OG0003380

TRINITY_DN28518_c0_g1_i1|m.64186_nitz2144 OG0003380

P. multiseries

LN865384.1_PsnmuV1.4_scaffold_220 7444-8199 -

LN865619.1_PsnmuV1.4_scaffold_455 25524-26258 -

73

Table 2-3. (Cont.) Species name Gene ID/ Scaffold name Approximate

gene location coordinates

Orthogroup ID

CMLE (3-carboxy-cis,cis-muconate lactonizing enzyme)

Nitzschia sp. Nitz4

maker-scaffold-212-augustus-gene-0.118 OG0005692

F. cylindrus

OEU21789.1 OG0005692

T. pseudonana

XP_02294624.1 OG0005692

P. tricornutum

XP_002186108.1 OG0005692

Nitzschia sp. Nitz2144

TRINITY_DN17337_c0_g1_i1|m.31016 OG0005692

P. multiseries

LN865398.1_PsnmuV1.4_scaffold_234-size_84702~1 10312 10944 -

CMD (4-carboxy-muconolactone decarboxylase)

Nitzschia sp. Nitz4

TRINITY_DN5181_c0_g1_i2|m.19823 -

TRINITY_DN5181_c0_g1_i3|m.19827 -

TRINITY_DN5181_c0_g1_i5|m.19835 -

TRINITY_DN5181_c0_g1_i9|m.19850 -

F. cylindrus

OEU17373.1 OG0007118

KV784448.1_FRACYscaffold_97 126002-126649 -

P. tricornutum

NC_011679.1_chromosome_11_complete_sequence 672175-672819 -

T. pseudonana

XP_002296700.1 -

XP_02286031.1 OG0007118

74

Table 2-3. (Cont.) Species name Gene ID/ Scaffold name Approximate

gene location coordinates

Orthogroup ID

Nitzschia sp. Nitz2144

TRINITY_DN14112_c0_g1_i1|m.22486 OG0007118

TRINITY_DN14112_c0_g1_i2|m.22488 OG0007118

P. multiseries

LN865211.1_PsnmuV1.4_scaffold_47 182326-183117 -

ELH (enol-lactone hydrolase)

Nitzschia sp. Nitz4

snap_masked-scaffold-381-processed-gene-0.9-mRNA-1 OG0003891

F. cylindrus

OEU08712.1 OG0009503

T. pseudonana

XP_002294595.1 OG0010576

P. tricornutum

0 0 -

Nitzschia sp. Nitz2144

TRINITY_DN18903_c0_g1_i1|m.37015 OG0007049

P. multiseries

0 0 -

TR ( 3-ketoadipate:succinyl-CoA transferase)

Nitzschia sp. Nitz4

augustus_masked-scaffold-630-processed-gene-0.27 OG0007884

F. cylindrus

0 0 -

75

Table 2-3. (Cont.) Species name Gene ID/ Scaffold name Approximate

gene location coordinates

Orthogroup ID

T. pseudonana

XP_002288066.1 OG0007884

P. tricornutum

XP_002184327.1 OG0007884

Nitzschia sp. Nitz2144

TRINITY_DN13259_c0_g1_i1|m.20451 OG0007884

P. multiseries

0 0 -

TH (3-ketoadypil-CoA thiolase)

Nitzschia sp. Nitz4

maker-agouti_scaf_70-snap-gene-0.94-mRNA-1 OG0005358

F. cylindrus

OEU20070.1 OG0005358

T. pseudonana

XP_002291557.1 OG0005358

P. tricornutum

XP_002180306.1 OG0005358

Nitzschia sp. Nitz2144

TRINITY_DN11969_c0_g1_i1|m.17323 OG0005358

P. multiseries

LN865426.1_PsnmuV1.4_scaffold_262 60224 61021 -

PcaK ( 4-hydroxybenzoate and protocatechuate transporter)

Nitzschia sp. Nitz4

augustus_masked-scaffold-375-processed-gene-0.7-mRNA-1 OG0003582

76

Table 2-3. (Cont.) Species name Gene ID/ Scaffold name Approximate

gene location coordinates

Orthogroup ID

F. cylindrus

OEU08398.1 OG0003582

OEU07405.1 OG0003476

T. pseudonana

XP_002288818.1 OG0003582

XP_002287244.1 no_ortho

XP_002294330.1 OG0003476

P. tricornutum

XP_002178175.1 OG0003582

XP_002177299.1 OG0003476

Nitzschia sp. Nitz2144

TRINITY_DN22222_c0_g2_i1|m.53129 OG0003582

P. multiseries

LN865266.1_PsnmuV1.4_scaffold_102 96577-97404 -

LN865522.1_PsnmuV1.4_scaffold_358 9025-9774 -

77

Figure 2-1. Protocatechuate dioxygenase genes localization on Nitzschia sp. Nitz4 scaffolds

78

Figure 2-2. Schematic annotation of protocatechuate branch of 𝛽-ketoadipate pathway in analyzed diatom species. Green arrow -gene present, red -gene was not found, blue -putative gene was found.

79

Figure S2-1. Statistics plots for original genome scaffolding for range of K-mers

80

Figure S2-2. Statistics plots for organellar reads-free genome scaffolding for range of K-mers

81

Figure S2-3. Comprehensive mitochondrial metabolic pathways of Nitzschia sp. Nitz4 derived from genome annotation

82

Figure S2-4. Comprehensive plastid metabolic pathways of Nitzschia sp. Nitz4 derived from genome annotation

83

Conclusions The goal of this study was to determine the number of losses of photosynthesis in diatoms, a

diverse photosynthetic lineage of eukaryotic algae. In the first chapter of my thesis, I showed that

photosynthesis was lost just one time. This finding reframes both our understanding of this

radical trophic shift but also helps shape future research priorities aimed at understanding the

evolutionary, ecological, and genomic contexts of the switch from autotrophy to heterotrophy in

diatoms. The second overall goal of my project was to use the nuclear genome sequence of a

nonphotosynthetic diatom species to characterize core carbon metabolism and identify the

sources of extracellular carbon used by these species. This work, described in chapter two of my

thesis, led to the discovery of a novel carbon metabolic pathway involved in the degradation of

lignin-derived aromatic compounds. If verified experimentally, this would show that

nonphotosynthetic diatoms are able to capitalize on one of the most abundant sources of carbon

on the planet, further cementing their place among the most intriguing and important contributors

to the global carbon cycle.

84